forked from aosp-mirror/kernel_msm
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Semaphore CM11 + kexec hard boot #2
Open
raden
wants to merge
23
commits into
stratosk:kk-4.4.1
Choose a base branch
from
raden:semaphore-cm11
base: kk-4.4.1
Could not load branches
Branch not found: {{ refName }}
Could not load tags
Nothing to show
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…o semaphore-cm11
…semaphore-cm11
…into semaphore-cm11
…o semaphore-cm11
…-mako-kernel into semaphore-cm11
…o semaphore-cm11
…o semaphore-cm11
…into semaphore-cm11
stratosk
pushed a commit
that referenced
this pull request
Mar 12, 2014
commit 6f2e9f0 upstream. Now when we set the group inode free count, we don't have a proper group lock so that multiple threads may decrease the inode free count at the same time. And e2fsck will complain something like: Free inodes count wrong for group #1 (1, counted=0). Fix? no Free inodes count wrong for group #2 (3, counted=0). Fix? no Directories count wrong for group #2 (780, counted=779). Fix? no Free inodes count wrong for group #3 (2272, counted=2273). Fix? no So this patch try to protect it with the ext4_lock_group. btw, it is found by xfstests test case 269 and the volume is mkfsed with the parameter "-O ^resize_inode,^uninit_bg,extent,meta_bg,flex_bg,ext_attr" and I have run it 100 times and the error in e2fsck doesn't show up again. Signed-off-by: Tao Ma <boyu.mt@taobao.com> Signed-off-by: "Theodore Ts'o" <tytso@mit.edu> Signed-off-by: Benjamin LaHaise <bcrl@kvack.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Mar 12, 2014
commit bd49940 upstream. As the initial domain we are able to search/map certain regions of memory to harvest configuration data. For all low-level we use ACPI tables - for interrupts we use exclusively ACPI _PRT (so DSDT) and MADT for INT_SRC_OVR. The SMP MP table is not used at all. As a matter of fact we do not even support machines that only have SMP MP but no ACPI tables. Lets follow how Moorestown does it and just disable searching for BIOS SMP tables. This also fixes an issue on HP Proliant BL680c G5 and DL380 G6: 9f->100 for 1:1 PTE Freeing 9f-100 pfn range: 97 pages freed 1-1 mapping on 9f->100 .. snip.. e820: BIOS-provided physical RAM map: Xen: [mem 0x0000000000000000-0x000000000009efff] usable Xen: [mem 0x000000000009f400-0x00000000000fffff] reserved Xen: [mem 0x0000000000100000-0x00000000cfd1dfff] usable .. snip.. Scan for SMP in [mem 0x00000000-0x000003ff] Scan for SMP in [mem 0x0009fc00-0x0009ffff] Scan for SMP in [mem 0x000f0000-0x000fffff] found SMP MP-table at [mem 0x000f4fa0-0x000f4faf] mapped at [ffff8800000f4fa0] (XEN) mm.c:908:d0 Error getting mfn 100 (pfn 5555555555555555) from L1 entry 0000000000100461 for l1e_owner=0, pg_owner=0 (XEN) mm.c:4995:d0 ptwr_emulate: could not get_page_from_l1e() BUG: unable to handle kernel NULL pointer dereference at (null) IP: [<ffffffff81ac07e2>] xen_set_pte_init+0x66/0x71 . snip.. Pid: 0, comm: swapper Not tainted 3.6.0-rc6upstream-00188-gb6fb969-dirty #2 HP ProLiant BL680c G5 .. snip.. Call Trace: [<ffffffff81ad31c6>] __early_ioremap+0x18a/0x248 [<ffffffff81624731>] ? printk+0x48/0x4a [<ffffffff81ad32ac>] early_ioremap+0x13/0x15 [<ffffffff81acc140>] get_mpc_size+0x2f/0x67 [<ffffffff81acc284>] smp_scan_config+0x10c/0x136 [<ffffffff81acc2e4>] default_find_smp_config+0x36/0x5a [<ffffffff81ac3085>] setup_arch+0x5b3/0xb5b [<ffffffff81624731>] ? printk+0x48/0x4a [<ffffffff81abca7f>] start_kernel+0x90/0x390 [<ffffffff81abc356>] x86_64_start_reservations+0x131/0x136 [<ffffffff81abfa83>] xen_start_kernel+0x65f/0x661 (XEN) Domain 0 crashed: 'noreboot' set - not rebooting. which is that ioremap would end up mapping 0xff using _PAGE_IOMAP (which is what early_ioremap sticks as a flag) - which meant we would get MFN 0xFF (pte ff461, which is OK), and then it would also map 0x100 (b/c ioremap tries to get page aligned request, and it was trying to map 0xf4fa0 + PAGE_SIZE - so it mapped the next page) as _PAGE_IOMAP. Since 0x100 is actually a RAM page, and the _PAGE_IOMAP bypasses the P2M lookup we would happily set the PTE to 1000461. Xen would deny the request since we do not have access to the Machine Frame Number (MFN) of 0x100. The P2M[0x100] is for example 0x80140. Fixes-Oracle-Bugzilla: https://bugzilla.oracle.com/bugzilla/show_bug.cgi?id=13665 Acked-by: Jan Beulich <jbeulich@suse.com> Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> [bwh: Backported to 3.2: adjust context] Signed-off-by: Ben Hutchings <ben@decadent.org.uk> Cc: Yijing Wang <wangyijing@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
…o semaphore-cm11
stratosk
pushed a commit
that referenced
this pull request
Apr 6, 2014
commit d25f06e upstream. vmxnet3's netpoll driver is incorrectly coded. It directly calls vmxnet3_do_poll, which is the driver internal napi poll routine. As the netpoll controller method doesn't block real napi polls in any way, there is a potential for race conditions in which the netpoll controller method and the napi poll method run concurrently. The result is data corruption causing panics such as this one recently observed: PID: 1371 TASK: ffff88023762caa0 CPU: 1 COMMAND: "rs:main Q:Reg" #0 [ffff88023abd5780] machine_kexec at ffffffff81038f3b #1 [ffff88023abd57e0] crash_kexec at ffffffff810c5d92 #2 [ffff88023abd58b0] oops_end at ffffffff8152b570 #3 [ffff88023abd58e0] die at ffffffff81010e0b aosp-mirror#4 [ffff88023abd5910] do_trap at ffffffff8152add4 aosp-mirror#5 [ffff88023abd5970] do_invalid_op at ffffffff8100cf95 aosp-mirror#6 [ffff88023abd5a10] invalid_op at ffffffff8100bf9b [exception RIP: vmxnet3_rq_rx_complete+1968] RIP: ffffffffa00f1e80 RSP: ffff88023abd5ac8 RFLAGS: 00010086 RAX: 0000000000000000 RBX: ffff88023b5dcee0 RCX: 00000000000000c0 RDX: 0000000000000000 RSI: 00000000000005f2 RDI: ffff88023b5dcee0 RBP: ffff88023abd5b48 R8: 0000000000000000 R9: ffff88023a3b6048 R10: 0000000000000000 R11: 0000000000000002 R12: ffff8802398d4cd8 R13: ffff88023af35140 R14: ffff88023b60c890 R15: 0000000000000000 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 aosp-mirror#7 [ffff88023abd5b50] vmxnet3_do_poll at ffffffffa00f204a [vmxnet3] aosp-mirror#8 [ffff88023abd5b80] vmxnet3_netpoll at ffffffffa00f209c [vmxnet3] aosp-mirror#9 [ffff88023abd5ba0] netpoll_poll_dev at ffffffff81472bb7 The fix is to do as other drivers do, and have the poll controller call the top half interrupt handler, which schedules a napi poll properly to recieve frames Tested by myself, successfully. Signed-off-by: Neil Horman <nhorman@tuxdriver.com> CC: Shreyas Bhatewara <sbhatewara@vmware.com> CC: "VMware, Inc." <pv-drivers@vmware.com> CC: "David S. Miller" <davem@davemloft.net> Reviewed-by: Shreyas N Bhatewara <sbhatewara@vmware.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
…o semaphore-cm11
Signed-off-by: Muhammad Najmi Ahmad Zabidi <najmi.zabidi@gmail.com>
stratosk
pushed a commit
that referenced
this pull request
Apr 16, 2014
commit fc108d2 upstream. Release the lock before mmc_signal_sdio_irq is called by mxs_mmc_enable_sdio_irq. Backtrace: [ 65.470000] ============================================= [ 65.470000] [ INFO: possible recursive locking detected ] [ 65.470000] 3.5.0-rc5 #2 Not tainted [ 65.470000] --------------------------------------------- [ 65.470000] ksdioirqd/mmc0/73 is trying to acquire lock: [ 65.470000] (&(&host->lock)->rlock#2){-.-...}, at: [<bf054120>] mxs_mmc_enable_sdio_irq+0x18/0xdc [mxs_mmc] [ 65.470000] [ 65.470000] but task is already holding lock: [ 65.470000] (&(&host->lock)->rlock#2){-.-...}, at: [<bf054120>] mxs_mmc_enable_sdio_irq+0x18/0xdc [mxs_mmc] [ 65.470000] [ 65.470000] other info that might help us debug this: [ 65.470000] Possible unsafe locking scenario: [ 65.470000] [ 65.470000] CPU0 [ 65.470000] ---- [ 65.470000] lock(&(&host->lock)->rlock#2); [ 65.470000] lock(&(&host->lock)->rlock#2); [ 65.470000] [ 65.470000] *** DEADLOCK *** [ 65.470000] [ 65.470000] May be due to missing lock nesting notation [ 65.470000] [ 65.470000] 1 lock held by ksdioirqd/mmc0/73: [ 65.470000] #0: (&(&host->lock)->rlock#2){-.-...}, at: [<bf054120>] mxs_mmc_enable_sdio_irq+0x18/0xdc [mxs_mmc] [ 65.470000] [ 65.470000] stack backtrace: [ 65.470000] [<c0014990>] (unwind_backtrace+0x0/0xf4) from [<c005ccb8>] (__lock_acquire+0x14f8/0x1b98) [ 65.470000] [<c005ccb8>] (__lock_acquire+0x14f8/0x1b98) from [<c005d3f8>] (lock_acquire+0xa0/0x108) [ 65.470000] [<c005d3f8>] (lock_acquire+0xa0/0x108) from [<c02f671c>] (_raw_spin_lock_irqsave+0x48/0x5c) [ 65.470000] [<c02f671c>] (_raw_spin_lock_irqsave+0x48/0x5c) from [<bf054120>] (mxs_mmc_enable_sdio_irq+0x18/0xdc [mxs_mmc]) [ 65.470000] [<bf054120>] (mxs_mmc_enable_sdio_irq+0x18/0xdc [mxs_mmc]) from [<bf0541d0>] (mxs_mmc_enable_sdio_irq+0xc8/0xdc [mxs_mmc]) [ 65.470000] [<bf0541d0>] (mxs_mmc_enable_sdio_irq+0xc8/0xdc [mxs_mmc]) from [<c0219b38>] (sdio_irq_thread+0x1bc/0x274) [ 65.470000] [<c0219b38>] (sdio_irq_thread+0x1bc/0x274) from [<c003c324>] (kthread+0x8c/0x98) [ 65.470000] [<c003c324>] (kthread+0x8c/0x98) from [<c00101ac>] (kernel_thread_exit+0x0/0x8) [ 65.470000] BUG: spinlock lockup suspected on CPU#0, ksdioirqd/mmc0/73 [ 65.470000] lock: 0xc3358724, .magic: dead4ead, .owner: ksdioirqd/mmc0/73, .owner_cpu: 0 [ 65.470000] [<c0014990>] (unwind_backtrace+0x0/0xf4) from [<c01b46b0>] (do_raw_spin_lock+0x100/0x144) [ 65.470000] [<c01b46b0>] (do_raw_spin_lock+0x100/0x144) from [<c02f6724>] (_raw_spin_lock_irqsave+0x50/0x5c) [ 65.470000] [<c02f6724>] (_raw_spin_lock_irqsave+0x50/0x5c) from [<bf054120>] (mxs_mmc_enable_sdio_irq+0x18/0xdc [mxs_mmc]) [ 65.470000] [<bf054120>] (mxs_mmc_enable_sdio_irq+0x18/0xdc [mxs_mmc]) from [<bf0541d0>] (mxs_mmc_enable_sdio_irq+0xc8/0xdc [mxs_mmc]) [ 65.470000] [<bf0541d0>] (mxs_mmc_enable_sdio_irq+0xc8/0xdc [mxs_mmc]) from [<c0219b38>] (sdio_irq_thread+0x1bc/0x274) [ 65.470000] [<c0219b38>] (sdio_irq_thread+0x1bc/0x274) from [<c003c324>] (kthread+0x8c/0x98) [ 65.470000] [<c003c324>] (kthread+0x8c/0x98) from [<c00101ac>] (kernel_thread_exit+0x0/0x8) Reported-by: Attila Kinali <attila@kinali.ch> Signed-off-by: Lauri Hintsala <lauri.hintsala@bluegiga.com> Acked-by: Shawn Guo <shawn.guo@linaro.org> Signed-off-by: Chris Ball <cjb@laptop.org> [bwh: Backported to 3.2: - Adjust context - HW_SSP_STATUS is a simple rather than function-like macro] Signed-off-by: Ben Hutchings <ben@decadent.org.uk> Cc: Qiang Huang <h.huangqiang@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
May 26, 2014
commit a585f87 upstream. The scenario here is that someone calls enable_irq_wake() from somewhere in the code. This will result in the lockdep producing a backtrace as can be seen below. In my case, this problem is triggered when using the wl1271 (TI WlCore) driver found in drivers/net/wireless/ti/ . The problem cause is rather obvious from the backtrace, but let's outline the dependency. enable_irq_wake() grabs the IRQ buslock in irq_set_irq_wake(), which in turns calls mxs_gpio_set_wake_irq() . But mxs_gpio_set_wake_irq() calls enable_irq_wake() again on the one-level-higher IRQ , thus it tries to grab the IRQ buslock again in irq_set_irq_wake() . Because the spinlock in irq_set_irq_wake()->irq_get_desc_buslock()->__irq_get_desc_lock() is not marked as recursive, lockdep will spew the stuff below. We know we can safely re-enter the lock, so use IRQ_GC_INIT_NESTED_LOCK to fix the spew. ============================================= [ INFO: possible recursive locking detected ] 3.10.33-00012-gf06b763-dirty #61 Not tainted --------------------------------------------- kworker/0:1/18 is trying to acquire lock: (&irq_desc_lock_class){-.-...}, at: [<c00685f0>] __irq_get_desc_lock+0x48/0x88 but task is already holding lock: (&irq_desc_lock_class){-.-...}, at: [<c00685f0>] __irq_get_desc_lock+0x48/0x88 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&irq_desc_lock_class); lock(&irq_desc_lock_class); *** DEADLOCK *** May be due to missing lock nesting notation 3 locks held by kworker/0:1/18: #0: (events){.+.+.+}, at: [<c0036308>] process_one_work+0x134/0x4a4 #1: ((&fw_work->work)){+.+.+.}, at: [<c0036308>] process_one_work+0x134/0x4a4 #2: (&irq_desc_lock_class){-.-...}, at: [<c00685f0>] __irq_get_desc_lock+0x48/0x88 stack backtrace: CPU: 0 PID: 18 Comm: kworker/0:1 Not tainted 3.10.33-00012-gf06b763-dirty #61 Workqueue: events request_firmware_work_func [<c0013eb4>] (unwind_backtrace+0x0/0xf0) from [<c0011c74>] (show_stack+0x10/0x14) [<c0011c74>] (show_stack+0x10/0x14) from [<c005bb08>] (__lock_acquire+0x140c/0x1a64) [<c005bb08>] (__lock_acquire+0x140c/0x1a64) from [<c005c6a8>] (lock_acquire+0x9c/0x104) [<c005c6a8>] (lock_acquire+0x9c/0x104) from [<c051d5a4>] (_raw_spin_lock_irqsave+0x44/0x58) [<c051d5a4>] (_raw_spin_lock_irqsave+0x44/0x58) from [<c00685f0>] (__irq_get_desc_lock+0x48/0x88) [<c00685f0>] (__irq_get_desc_lock+0x48/0x88) from [<c0068e78>] (irq_set_irq_wake+0x20/0xf4) [<c0068e78>] (irq_set_irq_wake+0x20/0xf4) from [<c027260c>] (mxs_gpio_set_wake_irq+0x1c/0x24) [<c027260c>] (mxs_gpio_set_wake_irq+0x1c/0x24) from [<c0068cf4>] (set_irq_wake_real+0x30/0x44) [<c0068cf4>] (set_irq_wake_real+0x30/0x44) from [<c0068ee4>] (irq_set_irq_wake+0x8c/0xf4) [<c0068ee4>] (irq_set_irq_wake+0x8c/0xf4) from [<c0310748>] (wlcore_nvs_cb+0x10c/0x97c) [<c0310748>] (wlcore_nvs_cb+0x10c/0x97c) from [<c02be5e8>] (request_firmware_work_func+0x38/0x58) [<c02be5e8>] (request_firmware_work_func+0x38/0x58) from [<c0036394>] (process_one_work+0x1c0/0x4a4) [<c0036394>] (process_one_work+0x1c0/0x4a4) from [<c0036a4c>] (worker_thread+0x138/0x394) [<c0036a4c>] (worker_thread+0x138/0x394) from [<c003cb74>] (kthread+0xa4/0xb0) [<c003cb74>] (kthread+0xa4/0xb0) from [<c000ee00>] (ret_from_fork+0x14/0x34) wlcore: loaded Signed-off-by: Marek Vasut <marex@denx.de> Acked-by: Shawn Guo <shawn.guo@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Jun 15, 2014
commit 085b7a4 upstream. layoutget's prepare hook can call rpc_exit with status = NFS4_OK (0). Because of this, nfs4_proc_layoutget can't depend on a 0 status to mean that the RPC was successfully sent, received and parsed. To fix this, use the result's len member to see if parsing took place. This fixes the following OOPS -- calling xdr_init_decode() with a buffer length 0 doesn't set the stream's 'p' member and ends up using uninitialized memory in filelayout_decode_layout. BUG: unable to handle kernel paging request at 0000000000008050 IP: [<ffffffff81282e78>] memcpy+0x18/0x120 PGD 0 Oops: 0000 [#1] SMP last sysfs file: /sys/devices/pci0000:00/0000:00:11.0/0000:02:01.0/irq CPU 1 Modules linked in: nfs_layout_nfsv41_files nfs lockd fscache auth_rpcgss nfs_acl autofs4 sunrpc ipt_REJECT nf_conntrack_ipv4 nf_defrag_ipv4 iptable_filter ip_tables ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables ipv6 dm_mirror dm_region_hash dm_log dm_mod ppdev parport_pc parport snd_ens1371 snd_rawmidi snd_ac97_codec ac97_bus snd_seq snd_seq_device snd_pcm snd_timer snd soundcore snd_page_alloc e1000 microcode vmware_balloon i2c_piix4 i2c_core sg shpchp ext4 mbcache jbd2 sr_mod cdrom sd_mod crc_t10dif pata_acpi ata_generic ata_piix mptspi mptscsih mptbase scsi_transport_spi [last unloaded: speedstep_lib] Pid: 1665, comm: flush-0:22 Not tainted 2.6.32-356-test-2 #2 VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform RIP: 0010:[<ffffffff81282e78>] [<ffffffff81282e78>] memcpy+0x18/0x120 RSP: 0018:ffff88003dfab588 EFLAGS: 00010206 RAX: ffff88003dc42000 RBX: ffff88003dfab610 RCX: 0000000000000009 RDX: 000000003f807ff0 RSI: 0000000000008050 RDI: ffff88003dc42000 RBP: ffff88003dfab5b0 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000080 R12: 0000000000000024 R13: ffff88003dc42000 R14: ffff88003f808030 R15: ffff88003dfab6a0 FS: 0000000000000000(0000) GS:ffff880003420000(0000) knlGS:0000000000000000 CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b CR2: 0000000000008050 CR3: 000000003bc92000 CR4: 00000000001407e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process flush-0:22 (pid: 1665, threadinfo ffff88003dfaa000, task ffff880037f77540) Stack: ffffffffa0398ac1 ffff8800397c5940 ffff88003dfab610 ffff88003dfab6a0 <d> ffff88003dfab5d0 ffff88003dfab680 ffffffffa01c150b ffffea0000d82e70 <d> 000000508116713b 0000000000000000 0000000000000000 0000000000000000 Call Trace: [<ffffffffa0398ac1>] ? xdr_inline_decode+0xb1/0x120 [sunrpc] [<ffffffffa01c150b>] filelayout_decode_layout+0xeb/0x350 [nfs_layout_nfsv41_files] [<ffffffffa01c17fc>] filelayout_alloc_lseg+0x8c/0x3c0 [nfs_layout_nfsv41_files] [<ffffffff8150e6ce>] ? __wait_on_bit+0x7e/0x90 Signed-off-by: Weston Andros Adamson <dros@netapp.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com> [bwh: Backported to 3.2: adjust context] Signed-off-by: Ben Hutchings <ben@decadent.org.uk> Cc: Rui Xiang <rui.xiang@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Aug 19, 2014
commit 504d587 upstream. clockevents_increase_min_delta() calls printk() from under hrtimer_bases.lock. That causes lock inversion on scheduler locks because printk() can call into the scheduler. Lockdep puts it as: ====================================================== [ INFO: possible circular locking dependency detected ] 3.15.0-rc8-06195-g939f04b #2 Not tainted ------------------------------------------------------- trinity-main/74 is trying to acquire lock: (&port_lock_key){-.....}, at: [<811c60be>] serial8250_console_write+0x8c/0x10c but task is already holding lock: (hrtimer_bases.lock){-.-...}, at: [<8103caeb>] hrtimer_try_to_cancel+0x13/0x66 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> aosp-mirror#5 (hrtimer_bases.lock){-.-...}: [<8104a942>] lock_acquire+0x92/0x101 [<8142f11d>] _raw_spin_lock_irqsave+0x2e/0x3e [<8103c918>] __hrtimer_start_range_ns+0x1c/0x197 [<8107ec20>] perf_swevent_start_hrtimer.part.41+0x7a/0x85 [<81080792>] task_clock_event_start+0x3a/0x3f [<810807a4>] task_clock_event_add+0xd/0x14 [<8108259a>] event_sched_in+0xb6/0x17a [<810826a2>] group_sched_in+0x44/0x122 [<81082885>] ctx_sched_in.isra.67+0x105/0x11f [<810828e6>] perf_event_sched_in.isra.70+0x47/0x4b [<81082bf6>] __perf_install_in_context+0x8b/0xa3 [<8107eb8e>] remote_function+0x12/0x2a [<8105f5af>] smp_call_function_single+0x2d/0x53 [<8107e17d>] task_function_call+0x30/0x36 [<8107fb82>] perf_install_in_context+0x87/0xbb [<810852c9>] SYSC_perf_event_open+0x5c6/0x701 [<810856f9>] SyS_perf_event_open+0x17/0x19 [<8142f8ee>] syscall_call+0x7/0xb -> aosp-mirror#4 (&ctx->lock){......}: [<8104a942>] lock_acquire+0x92/0x101 [<8142f04c>] _raw_spin_lock+0x21/0x30 [<81081df3>] __perf_event_task_sched_out+0x1dc/0x34f [<8142cacc>] __schedule+0x4c6/0x4cb [<8142cae0>] schedule+0xf/0x11 [<8142f9a6>] work_resched+0x5/0x30 -> #3 (&rq->lock){-.-.-.}: [<8104a942>] lock_acquire+0x92/0x101 [<8142f04c>] _raw_spin_lock+0x21/0x30 [<81040873>] __task_rq_lock+0x33/0x3a [<8104184c>] wake_up_new_task+0x25/0xc2 [<8102474b>] do_fork+0x15c/0x2a0 [<810248a9>] kernel_thread+0x1a/0x1f [<814232a2>] rest_init+0x1a/0x10e [<817af949>] start_kernel+0x303/0x308 [<817af2ab>] i386_start_kernel+0x79/0x7d -> #2 (&p->pi_lock){-.-...}: [<8104a942>] lock_acquire+0x92/0x101 [<8142f11d>] _raw_spin_lock_irqsave+0x2e/0x3e [<810413dd>] try_to_wake_up+0x1d/0xd6 [<810414cd>] default_wake_function+0xb/0xd [<810461f3>] __wake_up_common+0x39/0x59 [<81046346>] __wake_up+0x29/0x3b [<811b8733>] tty_wakeup+0x49/0x51 [<811c3568>] uart_write_wakeup+0x17/0x19 [<811c5dc1>] serial8250_tx_chars+0xbc/0xfb [<811c5f28>] serial8250_handle_irq+0x54/0x6a [<811c5f57>] serial8250_default_handle_irq+0x19/0x1c [<811c56d8>] serial8250_interrupt+0x38/0x9e [<810510e7>] handle_irq_event_percpu+0x5f/0x1e2 [<81051296>] handle_irq_event+0x2c/0x43 [<81052cee>] handle_level_irq+0x57/0x80 [<81002a72>] handle_irq+0x46/0x5c [<810027df>] do_IRQ+0x32/0x89 [<8143036e>] common_interrupt+0x2e/0x33 [<8142f23c>] _raw_spin_unlock_irqrestore+0x3f/0x49 [<811c25a4>] uart_start+0x2d/0x32 [<811c2c04>] uart_write+0xc7/0xd6 [<811bc6f6>] n_tty_write+0xb8/0x35e [<811b9beb>] tty_write+0x163/0x1e4 [<811b9cd9>] redirected_tty_write+0x6d/0x75 [<810b6ed6>] vfs_write+0x75/0xb0 [<810b7265>] SyS_write+0x44/0x77 [<8142f8ee>] syscall_call+0x7/0xb -> #1 (&tty->write_wait){-.....}: [<8104a942>] lock_acquire+0x92/0x101 [<8142f11d>] _raw_spin_lock_irqsave+0x2e/0x3e [<81046332>] __wake_up+0x15/0x3b [<811b8733>] tty_wakeup+0x49/0x51 [<811c3568>] uart_write_wakeup+0x17/0x19 [<811c5dc1>] serial8250_tx_chars+0xbc/0xfb [<811c5f28>] serial8250_handle_irq+0x54/0x6a [<811c5f57>] serial8250_default_handle_irq+0x19/0x1c [<811c56d8>] serial8250_interrupt+0x38/0x9e [<810510e7>] handle_irq_event_percpu+0x5f/0x1e2 [<81051296>] handle_irq_event+0x2c/0x43 [<81052cee>] handle_level_irq+0x57/0x80 [<81002a72>] handle_irq+0x46/0x5c [<810027df>] do_IRQ+0x32/0x89 [<8143036e>] common_interrupt+0x2e/0x33 [<8142f23c>] _raw_spin_unlock_irqrestore+0x3f/0x49 [<811c25a4>] uart_start+0x2d/0x32 [<811c2c04>] uart_write+0xc7/0xd6 [<811bc6f6>] n_tty_write+0xb8/0x35e [<811b9beb>] tty_write+0x163/0x1e4 [<811b9cd9>] redirected_tty_write+0x6d/0x75 [<810b6ed6>] vfs_write+0x75/0xb0 [<810b7265>] SyS_write+0x44/0x77 [<8142f8ee>] syscall_call+0x7/0xb -> #0 (&port_lock_key){-.....}: [<8104a62d>] __lock_acquire+0x9ea/0xc6d [<8104a942>] lock_acquire+0x92/0x101 [<8142f11d>] _raw_spin_lock_irqsave+0x2e/0x3e [<811c60be>] serial8250_console_write+0x8c/0x10c [<8104e402>] call_console_drivers.constprop.31+0x87/0x118 [<8104f5d5>] console_unlock+0x1d7/0x398 [<8104fb70>] vprintk_emit+0x3da/0x3e4 [<81425f76>] printk+0x17/0x19 [<8105bfa0>] clockevents_program_min_delta+0x104/0x116 [<8105c548>] clockevents_program_event+0xe7/0xf3 [<8105cc1c>] tick_program_event+0x1e/0x23 [<8103c43c>] hrtimer_force_reprogram+0x88/0x8f [<8103c49e>] __remove_hrtimer+0x5b/0x79 [<8103cb21>] hrtimer_try_to_cancel+0x49/0x66 [<8103cb4b>] hrtimer_cancel+0xd/0x18 [<8107f102>] perf_swevent_cancel_hrtimer.part.60+0x2b/0x30 [<81080705>] task_clock_event_stop+0x20/0x64 [<81080756>] task_clock_event_del+0xd/0xf [<81081350>] event_sched_out+0xab/0x11e [<810813e0>] group_sched_out+0x1d/0x66 [<81081682>] ctx_sched_out+0xaf/0xbf [<81081e04>] __perf_event_task_sched_out+0x1ed/0x34f [<8142cacc>] __schedule+0x4c6/0x4cb [<8142cae0>] schedule+0xf/0x11 [<8142f9a6>] work_resched+0x5/0x30 other info that might help us debug this: Chain exists of: &port_lock_key --> &ctx->lock --> hrtimer_bases.lock Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(hrtimer_bases.lock); lock(&ctx->lock); lock(hrtimer_bases.lock); lock(&port_lock_key); *** DEADLOCK *** 4 locks held by trinity-main/74: #0: (&rq->lock){-.-.-.}, at: [<8142c6f3>] __schedule+0xed/0x4cb #1: (&ctx->lock){......}, at: [<81081df3>] __perf_event_task_sched_out+0x1dc/0x34f #2: (hrtimer_bases.lock){-.-...}, at: [<8103caeb>] hrtimer_try_to_cancel+0x13/0x66 #3: (console_lock){+.+...}, at: [<8104fb5d>] vprintk_emit+0x3c7/0x3e4 stack backtrace: CPU: 0 PID: 74 Comm: trinity-main Not tainted 3.15.0-rc8-06195-g939f04b #2 00000000 81c3a310 8b995c14 81426f69 8b995c44 81425a99 8161f671 8161f570 8161f538 8161f559 8161f538 8b995c78 8b142bb0 00000004 8b142fdc 8b142bb0 8b995ca8 8104a62d 8b142fac 000016f2 81c3a310 00000001 00000001 00000003 Call Trace: [<81426f69>] dump_stack+0x16/0x18 [<81425a99>] print_circular_bug+0x18f/0x19c [<8104a62d>] __lock_acquire+0x9ea/0xc6d [<8104a942>] lock_acquire+0x92/0x101 [<811c60be>] ? serial8250_console_write+0x8c/0x10c [<811c6032>] ? wait_for_xmitr+0x76/0x76 [<8142f11d>] _raw_spin_lock_irqsave+0x2e/0x3e [<811c60be>] ? serial8250_console_write+0x8c/0x10c [<811c60be>] serial8250_console_write+0x8c/0x10c [<8104af87>] ? lock_release+0x191/0x223 [<811c6032>] ? wait_for_xmitr+0x76/0x76 [<8104e402>] call_console_drivers.constprop.31+0x87/0x118 [<8104f5d5>] console_unlock+0x1d7/0x398 [<8104fb70>] vprintk_emit+0x3da/0x3e4 [<81425f76>] printk+0x17/0x19 [<8105bfa0>] clockevents_program_min_delta+0x104/0x116 [<8105cc1c>] tick_program_event+0x1e/0x23 [<8103c43c>] hrtimer_force_reprogram+0x88/0x8f [<8103c49e>] __remove_hrtimer+0x5b/0x79 [<8103cb21>] hrtimer_try_to_cancel+0x49/0x66 [<8103cb4b>] hrtimer_cancel+0xd/0x18 [<8107f102>] perf_swevent_cancel_hrtimer.part.60+0x2b/0x30 [<81080705>] task_clock_event_stop+0x20/0x64 [<81080756>] task_clock_event_del+0xd/0xf [<81081350>] event_sched_out+0xab/0x11e [<810813e0>] group_sched_out+0x1d/0x66 [<81081682>] ctx_sched_out+0xaf/0xbf [<81081e04>] __perf_event_task_sched_out+0x1ed/0x34f [<8104416d>] ? __dequeue_entity+0x23/0x27 [<81044505>] ? pick_next_task_fair+0xb1/0x120 [<8142cacc>] __schedule+0x4c6/0x4cb [<81047574>] ? trace_hardirqs_off_caller+0xd7/0x108 [<810475b0>] ? trace_hardirqs_off+0xb/0xd [<81056346>] ? rcu_irq_exit+0x64/0x77 Fix the problem by using printk_deferred() which does not call into the scheduler. Reported-by: Fengguang Wu <fengguang.wu@intel.com> Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 20, 2016
msm_sat_enqueue() calls spin_lock() and msm_sat_dequeue() calls spin_lock_irqsave(). This leads to lockdep warnings about the same lock being taken in interrupts on and interrupts off context which can lead to a potential deadlock. ================================= [ INFO: inconsistent lock state ] 3.4.0+ #382 Tainted: G W --------------------------------- inconsistent {HARDIRQ-ON-W} -> {IN-HARDIRQ-W} usage. kworker/u:2/94 [HC1[1]:SC0[0]:HE0:SE1] takes: (&(&sat->lock)->rlock){?.+...}, at: [<c0348b80>] msm_sat_enqueue+0x20/0xb0 {HARDIRQ-ON-W} state was registered at: [<c00bd290>] __lock_acquire+0x664/0x8d8 [<c00bd690>] lock_acquire+0x18c/0x1e8 [<c06989a4>] _raw_spin_lock+0x38/0x48 [<c0348b80>] msm_sat_enqueue+0x20/0xb0 [<c0349338>] msm_slim_rxwq+0x278/0x42c [<c0349580>] msm_slim_rx_msgq_thread+0x94/0x1f8 [<c008e480>] kthread+0x90/0xa0 [<c000f438>] kernel_thread_exit+0x0/0x8 irq event stamp: 24219 hardirqs last enabled at (24218): [<c0699188>] _raw_spin_unlock_irqrestore+0x3c/0x68 hardirqs last disabled at (24219): [<c06993f4>] __irq_svc+0x34/0x78 softirqs last enabled at (24129): [<c00738a8>] irq_exit+0x54/0xa8 softirqs last disabled at (24108): [<c00738a8>] irq_exit+0x54/0xa8 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&(&sat->lock)->rlock); <Interrupt> lock(&(&sat->lock)->rlock); *** DEADLOCK *** 3 locks held by kworker/u:2/94: #0: ((sat->satcl.name)){.+.+.+}, at: [<c0087124>] process_one_work+0x1e8/0x648 stratosk#1: ((&sat->wd)){+.+.+.}, at: [<c0087124>] process_one_work+0x1e8/0x648 stratosk#2: (&dev->tx_lock){+.+...}, at: [<c0349cac>] msm_xfer_msg+0xb4/0x51c stack backtrace: [<c00151b0>] (unwind_backtrace+0x0/0x120) from [<c00b9d48>] (print_usage_bug+0x258/0x2c0) [<c00b9d48>] (print_usage_bug+0x258/0x2c0) from [<c00ba12c>] (mark_lock+0x37c/0x68c) [<c00ba12c>] (mark_lock+0x37c/0x68c) from [<c00bd20c>] (__lock_acquire+0x5e0/0x8d8) [<c00bd20c>] (__lock_acquire+0x5e0/0x8d8) from [<c00bd690>] (lock_acquire+0x18c/0x1e8) [<c00bd690>] (lock_acquire+0x18c/0x1e8) from [<c06989a4>] (_raw_spin_lock+0x38/0x48) [<c06989a4>] (_raw_spin_lock+0x38/0x48) from [<c0348b80>] (msm_sat_enqueue+0x20/0xb0) [<c0348b80>] (msm_sat_enqueue+0x20/0xb0) from [<c0348dc8>] (msm_slim_interrupt+0x1b8/0x4b0) [<c0348dc8>] (msm_slim_interrupt+0x1b8/0x4b0) from [<c00d2ccc>] (handle_irq_event_percpu+0x118/0x3a4) [<c00d2ccc>] (handle_irq_event_percpu+0x118/0x3a4) from [<c00d2f94>] (handle_irq_event+0x3c/0x5c) [<c00d2f94>] (handle_irq_event+0x3c/0x5c) from [<c00d5c4c>] (handle_fasteoi_irq+0xd0/0x11c) [<c00d5c4c>] (handle_fasteoi_irq+0xd0/0x11c) from [<c00d2ae4>] (generic_handle_irq+0x24/0x2c) [<c00d2ae4>] (generic_handle_irq+0x24/0x2c) from [<c000f370>] (handle_IRQ+0x7c/0xc0) [<c000f370>] (handle_IRQ+0x7c/0xc0) from [<c0008774>] (gic_handle_irq+0x6c/0xc4) [<c0008774>] (gic_handle_irq+0x6c/0xc4) from [<c0699404>] (__irq_svc+0x44/0x78) Exception stack(0xee061cd8 to 0xee061d20) 1cc0: 00000001 eebda7c8 1ce0: 00000000 eebda400 20000013 c12b5430 ec96c168 c12b542c c12b5430 00000001 1d00: 20000013 ee061e14 3eb13eb1 ee061d20 c00ba520 c069918c 20000013 ffffffff [<c0699404>] (__irq_svc+0x44/0x78) from [<c069918c>] (_raw_spin_unlock_irqrestore+0x40/0x68) [<c069918c>] (_raw_spin_unlock_irqrestore+0x40/0x68) from [<c0279f2c>] (__debug_object_init+0x30c/0x344) [<c0279f2c>] (__debug_object_init+0x30c/0x344) from [<c007b068>] (init_timer_on_stack_key+0x18/0x30) [<c007b068>] (init_timer_on_stack_key+0x18/0x30) from [<c0695248>] (schedule_timeout+0x8c/0x4c0) [<c0695248>] (schedule_timeout+0x8c/0x4c0) from [<c069768c>] (wait_for_common+0xec/0x164) [<c069768c>] (wait_for_common+0xec/0x164) from [<c0349fc4>] (msm_xfer_msg+0x3cc/0x51c) [<c0349fc4>] (msm_xfer_msg+0x3cc/0x51c) from [<c034a358>] (slim_sat_rxprocess+0x244/0x664) [<c034a358>] (slim_sat_rxprocess+0x244/0x664) from [<c0087290>] (process_one_work+0x354/0x648) [<c0087290>] (process_one_work+0x354/0x648) from [<c0089754>] (worker_thread+0x1a8/0x2a8) [<c0089754>] (worker_thread+0x1a8/0x2a8) from [<c008e480>] (kthread+0x90/0xa0) [<c008e480>] (kthread+0x90/0xa0) from [<c000f438>] (kernel_thread_exit+0x0/0x8) Make this lock irqsafe as well so that this potential bug doesn't occur. Change-Id: Icbef6d1d749ee6ee81b079e19e57f22c38f00c68 Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 20, 2016
workqueue: change BUG_ON() to WARN_ON() This BUG_ON() can be triggered if you call schedule_work() before calling INIT_WORK(). It is a bug definitely, but it's nicer to just print a stack trace and return. Reported-by: Matt Renzelmann <mjr@cs.wisc.edu> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: Catch more locking problems with flush_work() If a workqueue is flushed with flush_work() lockdep checking can be circumvented. For example: static DEFINE_MUTEX(mutex); static void my_work(struct work_struct *w) { mutex_lock(&mutex); mutex_unlock(&mutex); } static DECLARE_WORK(work, my_work); static int __init start_test_module(void) { schedule_work(&work); return 0; } module_init(start_test_module); static void __exit stop_test_module(void) { mutex_lock(&mutex); flush_work(&work); mutex_unlock(&mutex); } module_exit(stop_test_module); would not always print a warning when flush_work() was called. In this trivial example nothing could go wrong since we are guaranteed module_init() and module_exit() don't run concurrently, but if the work item is schedule asynchronously we could have a scenario where the work item is running just at the time flush_work() is called resulting in a classic ABBA locking problem. Add a lockdep hint by acquiring and releasing the work item lockdep_map in flush_work() so that we always catch this potential deadlock scenario. Signed-off-by: Stephen Boyd <sboyd@codeaurora.org> Reviewed-by: Yong Zhang <yong.zhang0@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> lockdep: fix oops in processing workqueue Under memory load, on x86_64, with lockdep enabled, the workqueue's process_one_work() has been seen to oops in __lock_acquire(), barfing on a 0xffffffff00000000 pointer in the lockdep_map's class_cache[]. Because it's permissible to free a work_struct from its callout function, the map used is an onstack copy of the map given in the work_struct: and that copy is made without any locking. Surprisingly, gcc (4.5.1 in Hugh's case) uses "rep movsl" rather than "rep movsq" for that structure copy: which might race with a workqueue user's wait_on_work() doing lock_map_acquire() on the source of the copy, putting a pointer into the class_cache[], but only in time for the top half of that pointer to be copied to the destination map. Boom when process_one_work() subsequently does lock_map_acquire() on its onstack copy of the lockdep_map. Fix this, and a similar instance in call_timer_fn(), with a lockdep_copy_map() function which additionally NULLs the class_cache[]. Note: this oops was actually seen on 3.4-next, where flush_work() newly does the racing lock_map_acquire(); but Tejun points out that 3.4 and earlier are already vulnerable to the same through wait_on_work(). * Patch orginally from Peter. Hugh modified it a bit and wrote the description. Signed-off-by: Peter Zijlstra <peterz@infradead.org> Reported-by: Hugh Dickins <hughd@google.com> LKML-Reference: <alpine.LSU.2.00.1205070951170.1544@eggly.anvils> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: perform cpu down operations from low priority cpu_notifier() Currently, all workqueue cpu hotplug operations run off CPU_PRI_WORKQUEUE which is higher than normal notifiers. This is to ensure that workqueue is up and running while bringing up a CPU before other notifiers try to use workqueue on the CPU. Per-cpu workqueues are supposed to remain working and bound to the CPU for normal CPU_DOWN_PREPARE notifiers. This holds mostly true even with workqueue offlining running with higher priority because workqueue CPU_DOWN_PREPARE only creates a bound trustee thread which runs the per-cpu workqueue without concurrency management without explicitly detaching the existing workers. However, if the trustee needs to create new workers, it creates unbound workers which may wander off to other CPUs while CPU_DOWN_PREPARE notifiers are in progress. Furthermore, if the CPU down is cancelled, the per-CPU workqueue may end up with workers which aren't bound to the CPU. While reliably reproducible with a convoluted artificial test-case involving scheduling and flushing CPU burning work items from CPU down notifiers, this isn't very likely to happen in the wild, and, even when it happens, the effects are likely to be hidden by the following successful CPU down. Fix it by using different priorities for up and down notifiers - high priority for up operations and low priority for down operations. Workqueue cpu hotplug operations will soon go through further cleanup. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: drop CPU_DYING notifier operation Workqueue used CPU_DYING notification to mark GCWQ_DISASSOCIATED. This was necessary because workqueue's CPU_DOWN_PREPARE happened before other DOWN_PREPARE notifiers and workqueue needed to stay associated across the rest of DOWN_PREPARE. After the previous patch, workqueue's DOWN_PREPARE happens after others and can set GCWQ_DISASSOCIATED directly. Drop CPU_DYING and let the trustee set GCWQ_DISASSOCIATED after disabling concurrency management. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: ROGUE workers are UNBOUND workers Currently, WORKER_UNBOUND is used to mark workers for the unbound global_cwq and WORKER_ROGUE is used to mark workers for disassociated per-cpu global_cwqs. Both are used to make the marked worker skip concurrency management and the only place they make any difference is in worker_enter_idle() where WORKER_ROGUE is used to skip scheduling idle timer, which can easily be replaced with trustee state testing. This patch replaces WORKER_ROGUE with WORKER_UNBOUND and drops WORKER_ROGUE. This is to prepare for removing trustee and handling disassociated global_cwqs as unbound. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: use mutex for global_cwq manager exclusion POOL_MANAGING_WORKERS is used to ensure that at most one worker takes the manager role at any given time on a given global_cwq. Trustee later hitched on it to assume manager adding blocking wait for the bit. As trustee already needed a custom wait mechanism, waiting for MANAGING_WORKERS was rolled into the same mechanism. Trustee is scheduled to be removed. This patch separates out MANAGING_WORKERS wait into per-pool mutex. Workers use mutex_trylock() to test for manager role and trustee uses mutex_lock() to claim manager roles. gcwq_claim/release_management() helpers are added to grab and release manager roles of all pools on a global_cwq. gcwq_claim_management() always grabs pool manager mutexes in ascending pool index order and uses pool index as lockdep subclass. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: drop @bind from create_worker() Currently, create_worker()'s callers are responsible for deciding whether the newly created worker should be bound to the associated CPU and create_worker() sets WORKER_UNBOUND only for the workers for the unbound global_cwq. Creation during normal operation is always via maybe_create_worker() and @bind is true. For workers created during hotplug, @bind is false. Normal operation path is planned to be used even while the CPU is going through hotplug operations or offline and this static decision won't work. Drop @bind from create_worker() and decide whether to bind by looking at GCWQ_DISASSOCIATED. create_worker() will also set WORKER_UNBOUND autmatically if disassociated. To avoid flipping GCWQ_DISASSOCIATED while create_worker() is in progress, the flag is now allowed to be changed only while holding all manager_mutexes on the global_cwq. This requires that GCWQ_DISASSOCIATED is not cleared behind trustee's back. CPU_ONLINE no longer clears DISASSOCIATED before flushing trustee, which clears DISASSOCIATED before rebinding remaining workers if asked to release. For cases where trustee isn't around, CPU_ONLINE clears DISASSOCIATED after flushing trustee. Also, now, first_idle has UNBOUND set on creation which is explicitly cleared by CPU_ONLINE while binding it. These convolutions will soon be removed by further simplification of CPU hotplug path. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: reimplement CPU online rebinding to handle idle workers Currently, if there are left workers when a CPU is being brough back online, the trustee kills all idle workers and scheduled rebind_work so that they re-bind to the CPU after the currently executing work is finished. This works for busy workers because concurrency management doesn't try to wake up them from scheduler callbacks, which require the target task to be on the local run queue. The busy worker bumps concurrency counter appropriately as it clears WORKER_UNBOUND from the rebind work item and it's bound to the CPU before returning to the idle state. To reduce CPU on/offlining overhead (as many embedded systems use it for powersaving) and simplify the code path, workqueue is planned to be modified to retain idle workers across CPU on/offlining. This patch reimplements CPU online rebinding such that it can also handle idle workers. As noted earlier, due to the local wakeup requirement, rebinding idle workers is tricky. All idle workers must be re-bound before scheduler callbacks are enabled. This is achieved by interlocking idle re-binding. Idle workers are requested to re-bind and then hold until all idle re-binding is complete so that no bound worker starts executing work item. Only after all idle workers are re-bound and parked, CPU_ONLINE proceeds to release them and queue rebind work item to busy workers thus guaranteeing scheduler callbacks aren't invoked until all idle workers are ready. worker_rebind_fn() is renamed to busy_worker_rebind_fn() and idle_worker_rebind() for idle workers is added. Rebinding logic is moved to rebind_workers() and now called from CPU_ONLINE after flushing trustee. While at it, add CPU sanity check in worker_thread(). Note that now a worker may become idle or the manager between trustee release and rebinding during CPU_ONLINE. As the previous patch updated create_worker() so that it can be used by regular manager while unbound and this patch implements idle re-binding, this is safe. This prepares for removal of trustee and keeping idle workers across CPU hotplugs. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: don't butcher idle workers on an offline CPU Currently, during CPU offlining, after all pending work items are drained, the trustee butchers all workers. Also, on CPU onlining failure, workqueue_cpu_callback() ensures that the first idle worker is destroyed. Combined, these guarantee that an offline CPU doesn't have any worker for it once all the lingering work items are finished. This guarantee isn't really necessary and makes CPU on/offlining more expensive than needs to be, especially for platforms which use CPU hotplug for powersaving. This patch lets offline CPUs removes idle worker butchering from the trustee and let a CPU which failed onlining keep the created first worker. The first worker is created if the CPU doesn't have any during CPU_DOWN_PREPARE and started right away. If onlining succeeds, the rebind_workers() call in CPU_ONLINE will rebind it like any other workers. If onlining fails, the worker is left alone till the next try. This makes CPU hotplugs cheaper by allowing global_cwqs to keep workers across them and simplifies code. Note that trustee doesn't re-arm idle timer when it's done and thus the disassociated global_cwq will keep all workers until it comes back online. This will be improved by further patches. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: remove CPU offline trustee With the previous changes, a disassociated global_cwq now can run as an unbound one on its own - it can create workers as necessary to drain remaining works after the CPU has been brought down and manage the number of workers using the usual idle timer mechanism making trustee completely redundant except for the actual unbinding operation. This patch removes the trustee and let a disassociated global_cwq manage itself. Unbinding is moved to a work item (for CPU affinity) which is scheduled and flushed from CPU_DONW_PREPARE. This patch moves nr_running clearing outside gcwq and manager locks to simplify the code. As nr_running is unused at the point, this is safe. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: simplify CPU hotplug code With trustee gone, CPU hotplug code can be simplified. * gcwq_claim/release_management() now grab and release gcwq lock too respectively and gained _and_lock and _and_unlock postfixes. * All CPU hotplug logic was implemented in workqueue_cpu_callback() which was called by workqueue_cpu_up/down_callback() for the correct priority. This was because up and down paths shared a lot of logic, which is no longer true. Remove workqueue_cpu_callback() and move all hotplug logic into the two actual callbacks. This patch doesn't make any functional changes. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: fix spurious CPU locality WARN from process_one_work() 25511a4776 "workqueue: reimplement CPU online rebinding to handle idle workers" added CPU locality sanity check in process_one_work(). It triggers if a worker is executing on a different CPU without UNBOUND or REBIND set. This works for all normal workers but rescuers can trigger this spuriously when they're serving the unbound or a disassociated global_cwq - rescuers don't have either flag set and thus its gcwq->cpu can be a different value including %WORK_CPU_UNBOUND. Fix it by additionally testing %GCWQ_DISASSOCIATED. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> LKML-Refence: <20120721213656.GA7783@linux.vnet.ibm.com> workqueue: reorder queueing functions so that _on() variants are on top Currently, queue/schedule[_delayed]_work_on() are located below the counterpart without the _on postifx even though the latter is usually implemented using the former. Swap them. This is cleanup and doesn't cause any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: make queueing functions return bool All queueing functions return 1 on success, 0 if the work item was already pending. Update them to return bool instead. This signifies better that they don't return 0 / -errno. This is cleanup and doesn't cause any functional difference. While at it, fix comment opening for schedule_work_on(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: add missing smp_wmb() in process_one_work() WORK_STRUCT_PENDING is used to claim ownership of a work item and process_one_work() releases it before starting execution. When someone else grabs PENDING, all pre-release updates to the work item should be visible and all updates made by the new owner should happen afterwards. Grabbing PENDING uses test_and_set_bit() and thus has a full barrier; however, clearing doesn't have a matching wmb. Given the preceding spin_unlock and use of clear_bit, I don't believe this can be a problem on an actual machine and there hasn't been any related report but it still is theretically possible for clear_pending to permeate upwards and happen before work->entry update. Add an explicit smp_wmb() before work_clear_pending(). Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: stable@vger.kernel.org workqueue: disable irq while manipulating PENDING Queueing operations use WORK_STRUCT_PENDING_BIT to synchronize access to the target work item. They first try to claim the bit and proceed with queueing only after that succeeds and there's a window between PENDING being set and the actual queueing where the task can be interrupted or preempted. There's also a similar window in process_one_work() when clearing PENDING. A work item is dequeued, gcwq->lock is released and then PENDING is cleared and the worker might get interrupted or preempted between releasing gcwq->lock and clearing PENDING. cancel[_delayed]_work_sync() tries to claim or steal PENDING. The function assumes that a work item with PENDING is either queued or in the process of being [de]queued. In the latter case, it busy-loops until either the work item loses PENDING or is queued. If canceling coincides with the above described interrupts or preemptions, the canceling task will busy-loop while the queueing or executing task is preempted. This patch keeps irq disabled across claiming PENDING and actual queueing and moves PENDING clearing in process_one_work() inside gcwq->lock so that busy looping from PENDING && !queued doesn't wait for interrupted/preempted tasks. Note that, in process_one_work(), setting last CPU and clearing PENDING got merged into single operation. This removes possible long busy-loops and will allow using try_to_grab_pending() from bh and irq contexts. v2: __queue_work() was testing preempt_count() to ensure that the caller has disabled preemption. This triggers spuriously if !CONFIG_PREEMPT_COUNT. Use preemptible() instead. Reported by Fengguang Wu. v3: Disable irq instead of preemption. IRQ will be disabled while grabbing gcwq->lock later anyway and this allows using try_to_grab_pending() from bh and irq contexts. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Fengguang Wu <fengguang.wu@intel.com> workqueue: set delayed_work->timer function on initialization delayed_work->timer.function is currently initialized during queue_delayed_work_on(). Export delayed_work_timer_fn() and set delayed_work timer function during delayed_work initialization together with other fields. This ensures the timer function is always valid on an initialized delayed_work. This is to help mod_delayed_work() implementation. To detect delayed_work users which diddle with the internal timer, trigger WARN if timer function doesn't match on queue. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: unify local CPU queueing handling Queueing functions have been using different methods to determine the local CPU. * queue_work() superflously uses get/put_cpu() to acquire and hold the local CPU across queue_work_on(). * delayed_work_timer_fn() uses smp_processor_id(). * queue_delayed_work() calls queue_delayed_work_on() with -1 @cpu which is interpreted as the local CPU. * flush_delayed_work[_sync]() were using raw_smp_processor_id(). * __queue_work() interprets %WORK_CPU_UNBOUND as local CPU if the target workqueue is bound one but nobody uses this. This patch converts all functions to uniformly use %WORK_CPU_UNBOUND to indicate local CPU and use the local binding feature of __queue_work(). unlikely() is dropped from %WORK_CPU_UNBOUND handling in __queue_work(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix zero @delay handling of queue_delayed_work_on() If @delay is zero and the dealyed_work is idle, queue_delayed_work() queues it for immediate execution; however, queue_delayed_work_on() lacks this logic and always goes through timer regardless of @delay. This patch moves 0 @delay handling logic from queue_delayed_work() to queue_delayed_work_on() so that both functions behave the same. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: move try_to_grab_pending() upwards try_to_grab_pending() will be used by to-be-implemented mod_delayed_work[_on](). Move try_to_grab_pending() and related functions above queueing functions. This patch only moves functions around. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: introduce WORK_OFFQ_FLAG_* Low WORK_STRUCT_FLAG_BITS bits of work_struct->data contain WORK_STRUCT_FLAG_* and flush color. If the work item is queued, the rest point to the cpu_workqueue with WORK_STRUCT_CWQ set; otherwise, WORK_STRUCT_CWQ is clear and the bits contain the last CPU number - either a real CPU number or one of WORK_CPU_*. Scheduled addition of mod_delayed_work[_on]() requires an additional flag, which is used only while a work item is off queue. There are more than enough bits to represent off-queue CPU number on both 32 and 64bits. This patch introduces WORK_OFFQ_FLAG_* which occupy the lower part of the @work->data high bits while off queue. This patch doesn't define any actual OFFQ flag yet. Off-queue CPU number is now shifted by WORK_OFFQ_CPU_SHIFT, which adds the number of bits used by OFFQ flags to WORK_STRUCT_FLAG_SHIFT, to make room for OFFQ flags. To avoid shift width warning with large WORK_OFFQ_FLAG_BITS, ulong cast is added to WORK_STRUCT_NO_CPU and, just in case, BUILD_BUG_ON() to check that there are enough bits to accomodate off-queue CPU number is added. This patch doesn't make any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: factor out __queue_delayed_work() from queue_delayed_work_on() This is to prepare for mod_delayed_work[_on]() and doesn't cause any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reorganize try_to_grab_pending() and __cancel_timer_work() * Use bool @is_dwork instead of @timer and let try_to_grab_pending() use to_delayed_work() to determine the delayed_work address. * Move timer handling from __cancel_work_timer() to try_to_grab_pending(). * Make try_to_grab_pending() use -EAGAIN instead of -1 for busy-looping and drop the ret local variable. * Add proper function comment to try_to_grab_pending(). This makes the code a bit easier to understand and will ease further changes. This patch doesn't make any functional change. v2: Use @is_dwork instead of @timer. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: mark a work item being canceled as such There can be two reasons try_to_grab_pending() can fail with -EAGAIN. One is when someone else is queueing or deqeueing the work item. With the previous patches, it is guaranteed that PENDING and queued state will soon agree making it safe to busy-retry in this case. The other is if multiple __cancel_work_timer() invocations are racing one another. __cancel_work_timer() grabs PENDING and then waits for running instances of the target work item on all CPUs while holding PENDING and !queued. try_to_grab_pending() invoked from another task will keep returning -EAGAIN while the current owner is waiting. Not distinguishing the two cases is okay because __cancel_work_timer() is the only user of try_to_grab_pending() and it invokes wait_on_work() whenever grabbing fails. For the first case, busy looping should be fine but wait_on_work() doesn't cause any critical problem. For the latter case, the new contender usually waits for the same condition as the current owner, so no unnecessarily extended busy-looping happens. Combined, these make __cancel_work_timer() technically correct even without irq protection while grabbing PENDING or distinguishing the two different cases. While the current code is technically correct, not distinguishing the two cases makes it difficult to use try_to_grab_pending() for other purposes than canceling because it's impossible to tell whether it's safe to busy-retry grabbing. This patch adds a mechanism to mark a work item being canceled. try_to_grab_pending() now disables irq on success and returns -EAGAIN to indicate that grabbing failed but PENDING and queued states are gonna agree soon and it's safe to busy-loop. It returns -ENOENT if the work item is being canceled and it may stay PENDING && !queued for arbitrary amount of time. __cancel_work_timer() is modified to mark the work canceling with WORK_OFFQ_CANCELING after grabbing PENDING, thus making try_to_grab_pending() fail with -ENOENT instead of -EAGAIN. Also, it invokes wait_on_work() iff grabbing failed with -ENOENT. This isn't necessary for correctness but makes it consistent with other future users of try_to_grab_pending(). v2: try_to_grab_pending() was testing preempt_count() to ensure that the caller has disabled preemption. This triggers spuriously if !CONFIG_PREEMPT_COUNT. Use preemptible() instead. Reported by Fengguang Wu. v3: Updated so that try_to_grab_pending() disables irq on success rather than requiring preemption disabled by the caller. This makes busy-looping easier and will allow try_to_grap_pending() to be used from bh/irq contexts. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Fengguang Wu <fengguang.wu@intel.com> workqueue: implement mod_delayed_work[_on]() Workqueue was lacking a mechanism to modify the timeout of an already pending delayed_work. delayed_work users have been working around this using several methods - using an explicit timer + work item, messing directly with delayed_work->timer, and canceling before re-queueing, all of which are error-prone and/or ugly. This patch implements mod_delayed_work[_on]() which behaves similarly to mod_timer() - if the delayed_work is idle, it's queued with the given delay; otherwise, its timeout is modified to the new value. Zero @delay guarantees immediate execution. v2: Updated to reflect try_to_grab_pending() changes. Now safe to be called from bh context. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> workqueue: fix CPU binding of flush_delayed_work[_sync]() delayed_work encodes the workqueue to use and the last CPU in delayed_work->work.data while it's on timer. The target CPU is implicitly recorded as the CPU the timer is queued on and delayed_work_timer_fn() queues delayed_work->work to the CPU it is running on. Unfortunately, this leaves flush_delayed_work[_sync]() no way to find out which CPU the delayed_work was queued for when they try to re-queue after killing the timer. Currently, it chooses the local CPU flush is running on. This can unexpectedly move a delayed_work queued on a specific CPU to another CPU and lead to subtle errors. There isn't much point in trying to save several bytes in struct delayed_work, which is already close to a hundred bytes on 64bit with all debug options turned off. This patch adds delayed_work->cpu to remember the CPU it's queued for. Note that if the timer is migrated during CPU down, the work item could be queued to the downed global_cwq after this change. As a detached global_cwq behaves like an unbound one, this doesn't change much for the delayed_work. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> workqueue: add missing wmb() in clear_work_data() Any operation which clears PENDING should be preceded by a wmb to guarantee that the next PENDING owner sees all the changes made before PENDING release. There are only two places where PENDING is cleared - set_work_cpu_and_clear_pending() and clear_work_data(). The caller of the former already does smp_wmb() but the latter doesn't have any. Move the wmb above set_work_cpu_and_clear_pending() into it and add one to clear_work_data(). There hasn't been any report related to this issue, and, given how clear_work_data() is used, it is extremely unlikely to have caused any actual problems on any architecture. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> workqueue: use enum value to set array size of pools in gcwq Commit 3270476a6c0ce322354df8679652f060d66526dc ('workqueue: reimplement WQ_HIGHPRI using a separate worker_pool') introduce separate worker_pool for HIGHPRI. Although there is NR_WORKER_POOLS enum value which represent size of pools, definition of worker_pool in gcwq doesn't use it. Using it makes code robust and prevent future mistakes. So change code to use this enum value. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: correct req_cpu in trace_workqueue_queue_work() When we do tracing workqueue_queue_work(), it records requested cpu. But, if !(@wq->flag & WQ_UNBOUND) and @cpu is WORK_CPU_UNBOUND, requested cpu is changed as local cpu. In case of @wq->flag & WQ_UNBOUND, above change is not occured, therefore it is reasonable to correct it. Use temporary local variable for storing requested cpu. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: change value of lcpu in __queue_delayed_work_on() We assign cpu id into work struct's data field in __queue_delayed_work_on(). In current implementation, when work is come in first time, current running cpu id is assigned. If we do __queue_delayed_work_on() with CPU A on CPU B, __queue_work() invoked in delayed_work_timer_fn() go into the following sub-optimal path in case of WQ_NON_REENTRANT. gcwq = get_gcwq(cpu); if (wq->flags & WQ_NON_REENTRANT && (last_gcwq = get_work_gcwq(work)) && last_gcwq != gcwq) { Change lcpu to @cpu and rechange lcpu to local cpu if lcpu is WORK_CPU_UNBOUND. It is sufficient to prevent to go into sub-optimal path. tj: Slightly rephrased the comment. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: introduce system_highpri_wq Commit 3270476a6c0ce322354df8679652f060d66526dc ('workqueue: reimplement WQ_HIGHPRI using a separate worker_pool') introduce separate worker pool for HIGHPRI. When we handle busyworkers for gcwq, it can be normal worker or highpri worker. But, we don't consider this difference in rebind_workers(), we use just system_wq for highpri worker. It makes mismatch between cwq->pool and worker->pool. It doesn't make error in current implementation, but possible in the future. Now, we introduce system_highpri_wq to use proper cwq for highpri workers in rebind_workers(). Following patch fix this issue properly. tj: Even apart from rebinding, having system_highpri_wq generally makes sense. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use system_highpri_wq for highpri workers in rebind_workers() In rebind_workers(), we do inserting a work to rebind to cpu for busy workers. Currently, in this case, we use only system_wq. This makes a possible error situation as there is mismatch between cwq->pool and worker->pool. To prevent this, we should use system_highpri_wq for highpri worker to match theses. This implements it. tj: Rephrased comment a bit. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use system_highpri_wq for unbind_work To speed cpu down processing up, use system_highpri_wq. As scheduling priority of workers on it is higher than system_wq and it is not contended by other normal works on this cpu, work on it is processed faster than system_wq. tj: CPU up/downs care quite a bit about latency these days. This shouldn't hurt anything and makes sense. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix checkpatch issues Fixed some checkpatch warnings. tj: adapted to wq/for-3.7 and massaged pr_xxx() format strings a bit. Signed-off-by: Valentin Ilie <valentin.ilie@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> LKML-Reference: <1345326762-21747-1-git-send-email-valentin.ilie@gmail.com> workqueue: make all workqueues non-reentrant By default, each per-cpu part of a bound workqueue operates separately and a work item may be executing concurrently on different CPUs. The behavior avoids some cross-cpu traffic but leads to subtle weirdities and not-so-subtle contortions in the API. * There's no sane usefulness in allowing a single work item to be executed concurrently on multiple CPUs. People just get the behavior unintentionally and get surprised after learning about it. Most either explicitly synchronize or use non-reentrant/ordered workqueue but this is error-prone. * flush_work() can't wait for multiple instances of the same work item on different CPUs. If a work item is executing on cpu0 and then queued on cpu1, flush_work() can only wait for the one on cpu1. Unfortunately, work items can easily cross CPU boundaries unintentionally when the queueing thread gets migrated. This means that if multiple queuers compete, flush_work() can't even guarantee that the instance queued right before it is finished before returning. * flush_work_sync() was added to work around some of the deficiencies of flush_work(). In addition to the usual flushing, it ensures that all currently executing instances are finished before returning. This operation is expensive as it has to walk all CPUs and at the same time fails to address competing queuer case. Incorrectly using flush_work() when flush_work_sync() is necessary is an easy error to make and can lead to bugs which are difficult to reproduce. * Similar problems exist for flush_delayed_work[_sync](). Other than the cross-cpu access concern, there's no benefit in allowing parallel execution and it's plain silly to have this level of contortion for workqueue which is widely used from core code to extremely obscure drivers. This patch makes all workqueues non-reentrant. If a work item is executing on a different CPU when queueing is requested, it is always queued to that CPU. This guarantees that any given work item can be executing on one CPU at maximum and if a work item is queued and executing, both are on the same CPU. The only behavior change which may affect workqueue users negatively is that non-reentrancy overrides the affinity specified by queue_work_on(). On a reentrant workqueue, the affinity specified by queue_work_on() is always followed. Now, if the work item is executing on one of the CPUs, the work item will be queued there regardless of the requested affinity. I've reviewed all workqueue users which request explicit affinity, and, fortunately, none seems to be crazy enough to exploit parallel execution of the same work item. This adds an additional busy_hash lookup if the work item was previously queued on a different CPU. This shouldn't be noticeable under any sane workload. Work item queueing isn't a very high-frequency operation and they don't jump across CPUs all the time. In a micro benchmark to exaggerate this difference - measuring the time it takes for two work items to repeatedly jump between two CPUs a number (10M) of times with busy_hash table densely populated, the difference was around 3%. While the overhead is measureable, it is only visible in pathological cases and the difference isn't huge. This change brings much needed sanity to workqueue and makes its behavior consistent with timer. I think this is the right tradeoff to make. This enables significant simplification of workqueue API. Simplification patches will follow. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: gut flush[_delayed]_work_sync() Now that all workqueues are non-reentrant, flush[_delayed]_work_sync() are equivalent to flush[_delayed]_work(). Drop the separate implementation and make them thin wrappers around flush[_delayed]_work(). * start_flush_work() no longer takes @wait_executing as the only left user - flush_work() - always sets it to %true. * __cancel_work_timer() uses flush_work() instead of wait_on_work(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: gut system_nrt[_freezable]_wq() Now that all workqueues are non-reentrant, system[_freezable]_wq() are equivalent to system_nrt[_freezable]_wq(). Replace the latter with wrappers around system[_freezable]_wq(). The wrapping goes through inline functions so that __deprecated can be added easily. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: cosmetic whitespace updates for macro definitions Consistently use the last tab position for '\' line continuation in complex macro definitions. This is to help the following patches. This patch is cosmetic. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use hotcpu_notifier() for workqueue_cpu_down_callback() workqueue_cpu_down_callback() is used only if HOTPLUG_CPU=y, so hotcpu_notifier() fits better than cpu_notifier(). When HOTPLUG_CPU=y, hotcpu_notifier() and cpu_notifier() are the same. When HOTPLUG_CPU=n, if we use cpu_notifier(), workqueue_cpu_down_callback() will be called during boot to do nothing, and the memory of workqueue_cpu_down_callback() and gcwq_unbind_fn() will be discarded after boot. If we use hotcpu_notifier(), we can avoid the no-op call of workqueue_cpu_down_callback() and the memory of workqueue_cpu_down_callback() and gcwq_unbind_fn() will be discard at build time: $ ls -l kernel/workqueue.o.cpu_notifier kernel/workqueue.o.hotcpu_notifier -rw-rw-r-- 1 laijs laijs 484080 Sep 15 11:31 kernel/workqueue.o.cpu_notifier -rw-rw-r-- 1 laijs laijs 478240 Sep 15 11:31 kernel/workqueue.o.hotcpu_notifier $ size kernel/workqueue.o.cpu_notifier kernel/workqueue.o.hotcpu_notifier text data bss dec hex filename 18513 2387 1221 22121 5669 kernel/workqueue.o.cpu_notifier 18082 2355 1221 21658 549a kernel/workqueue.o.hotcpu_notifier tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reimplement cancel_delayed_work() using try_to_grab_pending() cancel_delayed_work() can't be called from IRQ handlers due to its use of del_timer_sync() and can't cancel work items which are already transferred from timer to worklist. Also, unlike other flush and cancel functions, a canceled delayed_work would still point to the last associated cpu_workqueue. If the workqueue is destroyed afterwards and the work item is re-used on a different workqueue, the queueing code can oops trying to dereference already freed cpu_workqueue. This patch reimplements cancel_delayed_work() using try_to_grab_pending() and set_work_cpu_and_clear_pending(). This allows the function to be called from IRQ handlers and makes its behavior consistent with other flush / cancel functions. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> workqueue: UNBOUND -> REBIND morphing in rebind_workers() should be atomic The compiler may compile the following code into TWO write/modify instructions. worker->flags &= ~WORKER_UNBOUND; worker->flags |= WORKER_REBIND; so the other CPU may temporarily see worker->flags which doesn't have either WORKER_UNBOUND or WORKER_REBIND set and perform local wakeup prematurely. Fix it by using single explicit assignment via ACCESS_ONCE(). Because idle workers have another WORKER_NOT_RUNNING flag, this bug doesn't exist for them; however, update it to use the same pattern for consistency. tj: Applied the change to idle workers too and updated comments and patch description a bit. Change-Id: I9b95f51d146c40c31ba028668d6f412bd74c6026 Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org workqueue: move WORKER_REBIND clearing in rebind_workers() to the end of the function This doesn't make any functional difference and is purely to help the next patch to be simpler. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> workqueue: fix possible deadlock in idle worker rebinding Currently, rebind_workers() and idle_worker_rebind() are two-way interlocked. rebind_workers() waits for idle workers to finish rebinding and rebound idle workers wait for rebind_workers() to finish rebinding busy workers before proceeding. Unfortunately, this isn't enough. The second wait from idle workers is implemented as follows. wait_event(gcwq->rebind_hold, !(worker->flags & WORKER_REBIND)); rebind_workers() clears WORKER_REBIND, wakes up the idle workers and then returns. If CPU hotplug cycle happens again before one of the idle workers finishes the above wait_event(), rebind_workers() will repeat the first part of the handshake - set WORKER_REBIND again and wait for the idle worker to finish rebinding - and this leads to deadlock because the idle worker would be waiting for WORKER_REBIND to clear. This is fixed by adding another interlocking step at the end - rebind_workers() now waits for all the idle workers to finish the above WORKER_REBIND wait before returning. This ensures that all rebinding steps are complete on all idle workers before the next hotplug cycle can happen. This problem was diagnosed by Lai Jiangshan who also posted a patch to fix the issue, upon which this patch is based. This is the minimal fix and further patches are scheduled for the next merge window to simplify the CPU hotplug path. Signed-off-by: Tejun Heo <tj@kernel.org> Original-patch-by: Lai Jiangshan <laijs@cn.fujitsu.com> LKML-Reference: <1346516916-1991-3-git-send-email-laijs@cn.fujitsu.com> workqueue: restore POOL_MANAGING_WORKERS This patch restores POOL_MANAGING_WORKERS which was replaced by pool->manager_mutex by 6037315269 "workqueue: use mutex for global_cwq manager exclusion". There's a subtle idle worker depletion bug across CPU hotplug events and we need to distinguish an actual manager and CPU hotplug preventing management. POOL_MANAGING_WORKERS will be used for the former and manager_mutex the later. This patch just lays POOL_MANAGING_WORKERS on top of the existing manager_mutex and doesn't introduce any synchronization changes. The next patch will update it. Note that this patch fixes a non-critical anomaly where too_many_workers() may return %true spuriously while CPU hotplug is in progress. While the issue could schedule idle timer spuriously, it didn't trigger any actual misbehavior. tj: Rewrote patch description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix possible idle worker depletion across CPU hotplug To simplify both normal and CPU hotplug paths, worker management is prevented while CPU hoplug is in progress. This is achieved by CPU hotplug holding the same exclusion mechanism used by workers to ensure there's only one manager per pool. If someone else seems to be performing the manager role, workers proceed to execute work items. CPU hotplug using the same mechanism can lead to idle worker depletion because all workers could proceed to execute work items while CPU hotplug is in progress and CPU hotplug itself wouldn't actually perform the worker management duty - it doesn't guarantee that there's an idle worker left when it releases management. This idle worker depletion, under extreme circumstances, can break forward-progress guarantee and thus lead to deadlock. This patch fixes the bug by using separate mechanisms for manager exclusion among workers and hotplug exclusion. For manager exclusion, POOL_MANAGING_WORKERS which was restored by the previous patch is used. pool->manager_mutex is now only used for exclusion between the elected manager and CPU hotplug. The elected manager won't proceed without holding pool->manager_mutex. This ensures that the worker which won the manager position can't skip managing while CPU hotplug is in progress. It will block on manager_mutex and perform management after CPU hotplug is complete. Note that hotplug may happen while waiting for manager_mutex. A manager isn't either on idle or busy list and thus the hoplug code can't unbind/rebind it. Make the manager handle its own un/rebinding. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: always clear WORKER_REBIND in busy_worker_rebind_fn() busy_worker_rebind_fn() didn't clear WORKER_REBIND if rebinding failed (CPU is down again). This used to be okay because the flag wasn't used for anything else. However, after 25511a477 "workqueue: reimplement CPU online rebinding to handle idle workers", WORKER_REBIND is also used to command idle workers to rebind. If not cleared, the worker may confuse the next CPU_UP cycle by having REBIND spuriously set or oops / get stuck by prematurely calling idle_worker_rebind(). WARNING: at /work/os/wq/kernel/workqueue.c:1323 worker_thread+0x4cd/0x5 00() Hardware name: Bochs Modules linked in: test_wq(O-) Pid: 33, comm: kworker/1:1 Tainted: G O 3.6.0-rc1-work+ #3 Call Trace: [<ffffffff8109039f>] warn_slowpath_common+0x7f/0xc0 [<ffffffff810903fa>] warn_slowpath_null+0x1a/0x20 [<ffffffff810b3f1d>] worker_thread+0x4cd/0x500 [<ffffffff810bc16e>] kthread+0xbe/0xd0 [<ffffffff81bd2664>] kernel_thread_helper+0x4/0x10 ---[ end trace e977cf20f4661968 ]--- BUG: unable to handle kernel NULL pointer dereference at (null) IP: [<ffffffff810b3db0>] worker_thread+0x360/0x500 PGD 0 Oops: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC Modules linked in: test_wq(O-) CPU 0 Pid: 33, comm: kworker/1:1 Tainted: G W O 3.6.0-rc1-work+ #3 Bochs Bochs RIP: 0010:[<ffffffff810b3db0>] [<ffffffff810b3db0>] worker_thread+0x360/0x500 RSP: 0018:ffff88001e1c9de0 EFLAGS: 00010086 RAX: 0000000000000000 RBX: ffff88001e633e00 RCX: 0000000000004140 RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000009 RBP: ffff88001e1c9ea0 R08: 0000000000000000 R09: 0000000000000001 R10: 0000000000000002 R11: 0000000000000000 R12: ffff88001fc8d580 R13: ffff88001fc8d590 R14: ffff88001e633e20 R15: ffff88001e1c6900 FS: 0000000000000000(0000) GS:ffff88001fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 0000000000000000 CR3: 00000000130e8000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process kworker/1:1 (pid: 33, threadinfo ffff88001e1c8000, task ffff88001e1c6900) Stack: ffff880000000000 ffff88001e1c9e40 0000000000000001 ffff88001e1c8010 ffff88001e519c78 ffff88001e1c9e58 ffff88001e1c6900 ffff88001e1c6900 ffff88001e1c6900 ffff88001e1c6900 ffff88001fc8d340 ffff88001fc8d340 Call Trace: [<ffffffff810bc16e>] kthread+0xbe/0xd0 [<ffffffff81bd2664>] kernel_thread_helper+0x4/0x10 Code: b1 00 f6 43 48 02 0f 85 91 01 00 00 48 8b 43 38 48 89 df 48 8b 00 48 89 45 90 e8 ac f0 ff ff 3c 01 0f 85 60 01 00 00 48 8b 53 50 <8b> 02 83 e8 01 85 c0 89 02 0f 84 3b 01 00 00 48 8b 43 38 48 8b RIP [<ffffffff810b3db0>] worker_thread+0x360/0x500 RSP <ffff88001e1c9de0> CR2: 0000000000000000 There was no reason to keep WORKER_REBIND on failure in the first place - WORKER_UNBOUND is guaranteed to be set in such cases preventing incorrectly activating concurrency management. Always clear WORKER_REBIND. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reimplement idle worker rebinding Currently rebind_workers() uses rebinds idle workers synchronously before proceeding to requesting busy workers to rebind. This is necessary because all workers on @worker_pool->idle_list must be bound before concurrency management local wake-ups from the busy workers take place. Unfortunately, the synchronous idle rebinding is quite complicated. This patch reimplements idle rebinding to simplify the code path. Rather than trying to make all idle workers bound before rebinding busy workers, we simply remove all to-be-bound idle workers from the idle list and let them add themselves back after completing rebinding (successful or not). As only workers which finished rebinding can on on the idle worker list, the idle worker list is guaranteed to have only bound workers unless CPU went down again and local wake-ups are safe. After the change, @worker_pool->nr_idle may deviate than the actual number of idle workers on @worker_pool->idle_list. More specifically, nr_idle may be non-zero while ->idle_list is empty. All users of ->nr_idle and ->idle_list are audited. The only affected one is too_many_workers() which is updated to check %false if ->idle_list is empty regardless of ->nr_idle. After this patch, rebind_workers() no longer performs the nasty idle-rebind retries which require temporary release of gcwq->lock, and both unbinding and rebinding are atomic w.r.t. global_cwq->lock. worker->idle_rebind and global_cwq->rebind_hold are now unnecessary and removed along with the definition of struct idle_rebind. Changed from V1: 1) remove unlikely from too_many_workers(), ->idle_list can be empty anytime, even before this patch, no reason to use unlikely. 2) fix a small rebasing mistake. (which is from rebasing the orignal fixing patch to for-next) 3) add a lot of comments. 4) clear WORKER_REBIND unconditionaly in idle_worker_rebind() tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: WORKER_REBIND is no longer necessary for busy rebinding Because the old unbind/rebinding implementation wasn't atomic w.r.t. GCWQ_DISASSOCIATED manipulation which is protected by global_cwq->lock, we had to use two flags, WORKER_UNBOUND and WORKER_REBIND, to avoid incorrectly losing all NOT_RUNNING bits with back-to-back CPU hotplug operations; otherwise, completion of rebinding while another unbinding is in progress could clear UNBIND prematurely. Now that both unbind/rebinding are atomic w.r.t. GCWQ_DISASSOCIATED, there's no need to use two flags. Just one is enough. Don't use WORKER_REBIND for busy rebinding. tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: WORKER_REBIND is no longer necessary for idle rebinding Now both worker destruction and idle rebinding remove the worker from idle list while it's still idle, so list_empty(&worker->entry) can be used to test whether either is pending and WORKER_DIE to distinguish between the two instead making WORKER_REBIND unnecessary. Use list_empty(&worker->entry) to determine whether destruction or rebinding is pending. This simplifies worker state transitions. WORKER_REBIND is not needed anymore. Remove it. tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: rename manager_mutex to assoc_mutex Now that manager_mutex's role has changed from synchronizing manager role to excluding hotplug against manager, the name is misleading. As it is protecting the CPU-association of the gcwq now, rename it to assoc_mutex. This patch is pure rename and doesn't introduce any functional change. tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use __cpuinit instead of __devinit for cpu callbacks For workqueue hotplug callbacks, it makes less sense to use __devinit which discards the memory after boot if !HOTPLUG. __cpuinit, which discards the memory after boot if !HOTPLUG_CPU fits better. tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix possible stall on try_to_grab_pending() of a delayed work item Currently, when try_to_grab_pending() grabs a delayed work item, it leaves its linked work items alone on the delayed_works. The linked work items are always NO_COLOR and will cause future cwq_activate_first_delayed() increase cwq->nr_active incorrectly, and may cause the whole cwq to stall. For example, state: cwq->max_active = 1, cwq->nr_active = 1 one work in cwq->pool, many in cwq->delayed_works. step1: try_to_grab_pending() removes a work item from delayed_works but leaves its NO_COLOR linked work items on it. step2: Later on, cwq_activate_first_delayed() activates the linked work item increasing ->nr_active. step3: cwq->nr_active = 1, but all activated work items of the cwq are NO_COLOR. When they finish, cwq->nr_active will not be decreased due to NO_COLOR, and no further work items will be activated from cwq->delayed_works. the cwq stalls. Fix it by ensuring the target work item is activated before stealing PENDING in try_to_grab_pending(). This ensures that all the linked work items are activated without incorrectly bumping cwq->nr_active. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@kernel.org workqueue: reimplement work_on_cpu() using system_wq The existing work_on_cpu() implementation is hugely inefficient. It creates a new kthread, execute that single function and then let the kthread die on each invocation. Now that system_wq can handle concurrent executions, there's no advantage of doing this. Reimplement work_on_cpu() using system_wq which makes it simpler and way more efficient. stable: While this isn't a fix in itself, it's needed to fix a workqueue related bug in cpufreq/powernow-k8. AFAICS, this shouldn't break other existing users. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Jiri Kosina <jkosina@suse.cz> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Bjorn Helgaas <bhelgaas@google.com> Cc: Len Brown <lenb@kernel.org> Cc: Rafael J. Wysocki <rjw@sisk.pl> Cc: stable@vger.kernel.org workqueue: introduce cwq_set_max_active() helper for thaw_workqueues() Using a helper instead of open code makes thaw_workqueues() clearer. The helper will also be used by the next patch. tj: Slight update to comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use cwq_set_max_active() helper for workqueue_set_max_active() workqueue_set_max_active() may increase ->max_active without activating delayed works and may make the activation order differ from the queueing order. Both aren't strictly bugs but the resulting behavior could be a bit odd. To make things more consistent, use cwq_set_max_active() helper which immediately makes use of the newly increased max_mactive if there are delayed work items and also keeps the activation order. tj: Slight update to description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: remove spurious WARN_ON_ONCE(in_irq()) from try_to_grab_pending() e0aecdd874 ("workqueue: use irqsafe timer for delayed_work") made try_to_grab_pending() safe to use from irq context but forgot to remove WARN_ON_ONCE(in_irq()). Remove it. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Fengguang Wu <fengguang.wu@intel.com> workqueue: cancel_delayed_work() should return %false if work item is idle 57b30ae77b ("workqueue: reimplement cancel_delayed_work() using try_to_grab_pending()") made cancel_delayed_work() always return %true unless someone else is also trying to cancel the work item, which is broken - if the target work item is idle, the return value should be %false. try_to_grab_pending() indicates that the target work item was idle by zero return value. Use it for return. Note that this brings cancel_delayed_work() in line with __cancel_work_timer() in return value handling. Signed-off-by: Dan Magenheimer <dan.magenheimer@oracle.com> Signed-off-by: Tejun Heo <tj@kernel.org> LKML-Reference: <444a6439-b1a4-4740-9e7e-bc37267cfe73@default> workqueue: exit rescuer_thread() as TASK_RUNNING A rescue thread exiting TASK_INTERRUPTIBLE can lead to a task scheduling off, never to be seen again. In the case where this occurred, an exiting thread hit reiserfs homebrew conditional resched while holding a mutex, bringing the box to its knees. PID: 18105 TASK: ffff8807fd412180 CPU: 5 COMMAND: "kdmflush" #0 [ffff8808157e7670] schedule at ffffffff8143f489 #1 [ffff8808157e77b8] reiserfs_get_block at ffffffffa038ab2d [reiserfs] #2 [ffff8808157e79a8] __block_write_begin at ffffffff8117fb14 #3 [ffff8808157e7a98] reiserfs_write_begin at ffffffffa0388695 [reiserfs] #4 [ffff8808157e7ad8] generic_perform_write at ffffffff810ee9e2 #5 [ffff8808157e7b58] generic_file_buffered_write at ffffffff810eeb41 #6 [ffff8808157e7ba8] __generic_file_aio_write at ffffffff810f1a3a #7 [ffff8808157e7c58] generic_file_aio_write at ffffffff810f1c88 #8 [ffff8808157e7cc8] do_sync_write at ffffffff8114f850 #9 [ffff8808157e7dd8] do_acct_process at ffffffff810a268f [exception RIP: kernel_thread_helper] RIP: ffffffff8144a5c0 RSP: ffff8808157e7f58 RFLAGS: 00000202 RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffffffff8107af60 RDI: ffff8803ee491d18 RBP: 0000000000000000 R8: 0000000000000000 R9: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 Signed-off-by: Mike Galbraith <mgalbraith@suse.de> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org workqueue: mod_delayed_work_on() shouldn't queue timer on 0 delay 8376fe22c7 ("workqueue: implement mod_delayed_work[_on]()") implemented mod_delayed_work[_on]() using the improved try_to_grab_pending(). The function is later used, among others, to replace [__]candel_delayed_work() + queue_delayed_work() combinations. Unfortunately, a delayed_work item w/ zero @delay is handled slightly differently by mod_delayed_work_on() compared to queue_delayed_work_on(). The latter skips timer altogether and directly queues it using queue_work_on() while the former schedules timer which will expire on the closest tick. This means, when @delay is zero, that [__]cancel_delayed_work() + queue_delayed_work_on() makes the target item immediately executable while mod_delayed_work_on() may induce delay of upto a full tick. This somewhat subtle difference breaks some of the converted users. e.g. block queue plugging uses delayed_work for deferred processing and uses mod_delayed_work_on() when the queue needs to be immediately unplugged. The above problem manifested as noticeably higher number of context switches under certain circumstances. The difference in behavior was caused by missing special case handling for 0 delay in mod_delayed_work_on() compared to queue_delayed_work_on(). Joonsoo Kim posted a patch to add it - ("workqueue: optimize mod_delayed_work_on() when @delay == 0")[1]. The patch was queued for 3.8 but it was described as optimization and I missed that it was a correctness issue. As both queue_delayed_work_on() and mod_delayed_work_on() use __queue_delayed_work() for queueing, it seems that the better approach is to move the 0 delay special handling to the function instead of duplicating it in mod_delayed_work_on(). Fix the problem by moving 0 delay special case handling from queue_delayed_work_on() to __queue_delayed_work(). This replaces Joonsoo's patch. [1] http://thread.gmane.org/gmane.linux.kernel/1379011/focus=1379012 Signed-off-by: Tejun Heo <tj@kernel.org> Reported-and-tested-by: Anders Kaseorg <andersk@MIT.EDU> Reported-and-tested-by: Zlatko Calusic <zlatko.calusic@iskon.hr> LKML-Reference: <alpine.DEB.2.00.1211280953350.26602@dr-wily.mit.edu> LKML-Reference: <50A78AA9.5040904@iskon.hr> Cc: Joonsoo Kim <js1304@gmail.com> workqueue: trivial fix for return statement in work_busy() Return type of work_busy() is unsigned int. There is return statement returning boolean value, 'false' in work_busy(). It is not problem, because 'false' may be treated '0'. However, fixing it would make code robust. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: add WARN_ON_ONCE() on CPU number to wq_worker_waking_up() Recently, workqueue code has gone through some changes and we found some bugs related to concurrency management operations happening on the wrong CPU. When a worker is concurrency managed (!WORKER_NOT_RUNNIG), it should be bound to its associated cpu and woken up to that cpu. Add WARN_ON_ONCE() to verify this. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: convert BUG_ON()s in __queue_delayed_work() to WARN_ON_ONCE()s 8852aac25e ("workqueue: mod_delayed_work_on() shouldn't queue timer on 0 delay") unexpectedly uncovered a very nasty abuse of delayed_work in megaraid - it allocated work_struct, casted it to delayed_work and then pass that into queue_delayed_work(). Previously, this was okay because 0 @delay short-circuited to queue_work() before doing anything with delayed_work. 8852aac25e moved 0 @delay test into __queue_delayed_work() after sanity check on delayed_work making megaraid trigger BUG_ON(). Although megaraid is already fixed by c1d390d8e6 ("megaraid: fix BUG_ON() from incorrect use of delayed work"), this patch converts BUG_ON()s in __queue_delayed_work() to WARN_ON_ONCE()s so that such abusers, if there are more, trigger warning but don't crash the machine. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Xiaotian Feng <xtfeng@gmail.com> wq Change-Id: Ia3c507777a995f32bf6b40dc8318203e53134229 Signed-off-by: franciscofranco <franciscofranco.1990@gmail.com>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 20, 2016
This moves ARM over to the asm-generic/unaligned.h header. This has the benefit of better code generated especially for ARMv7 on gcc 4.7+ compilers. As Arnd Bergmann, points out: The asm-generic version uses the "struct" version for native-endian unaligned access and the "byteshift" version for the opposite endianess. The current ARM version however uses the "byteshift" implementation for both. Thanks to Nicolas Pitre for the excellent analysis: Test case: int foo (int *x) { return get_unaligned(x); } long long bar (long long *x) { return get_unaligned(x); } With the current ARM version: foo: ldrb r3, [r0, stratosk#2] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 2B], MEM[(const u8 *)x_1(D) + 2B] ldrb r1, [r0, stratosk#1] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 1B], MEM[(const u8 *)x_1(D) + 1B] ldrb r2, [r0, #0] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D)], MEM[(const u8 *)x_1(D)] mov r3, r3, asl #16 @ tmp154, MEM[(const u8 *)x_1(D) + 2B], ldrb r0, [r0, stratosk#3] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 3B], MEM[(const u8 *)x_1(D) + 3B] orr r3, r3, r1, asl aosp-mirror#8 @, tmp155, tmp154, MEM[(const u8 *)x_1(D) + 1B], orr r3, r3, r2 @ tmp157, tmp155, MEM[(const u8 *)x_1(D)] orr r0, r3, r0, asl #24 @,, tmp157, MEM[(const u8 *)x_1(D) + 3B], bx lr @ bar: stmfd sp!, {r4, r5, r6, r7} @, mov r2, #0 @ tmp184, ldrb r5, [r0, aosp-mirror#6] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 6B], MEM[(const u8 *)x_1(D) + 6B] ldrb r4, [r0, aosp-mirror#5] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 5B], MEM[(const u8 *)x_1(D) + 5B] ldrb ip, [r0, stratosk#2] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 2B], MEM[(const u8 *)x_1(D) + 2B] ldrb r1, [r0, aosp-mirror#4] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 4B], MEM[(const u8 *)x_1(D) + 4B] mov r5, r5, asl #16 @ tmp175, MEM[(const u8 *)x_1(D) + 6B], ldrb r7, [r0, stratosk#1] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 1B], MEM[(const u8 *)x_1(D) + 1B] orr r5, r5, r4, asl aosp-mirror#8 @, tmp176, tmp175, MEM[(const u8 *)x_1(D) + 5B], ldrb r6, [r0, aosp-mirror#7] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 7B], MEM[(const u8 *)x_1(D) + 7B] orr r5, r5, r1 @ tmp178, tmp176, MEM[(const u8 *)x_1(D) + 4B] ldrb r4, [r0, #0] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D)], MEM[(const u8 *)x_1(D)] mov ip, ip, asl #16 @ tmp188, MEM[(const u8 *)x_1(D) + 2B], ldrb r1, [r0, stratosk#3] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 3B], MEM[(const u8 *)x_1(D) + 3B] orr ip, ip, r7, asl aosp-mirror#8 @, tmp189, tmp188, MEM[(const u8 *)x_1(D) + 1B], orr r3, r5, r6, asl #24 @,, tmp178, MEM[(const u8 *)x_1(D) + 7B], orr ip, ip, r4 @ tmp191, tmp189, MEM[(const u8 *)x_1(D)] orr ip, ip, r1, asl #24 @, tmp194, tmp191, MEM[(const u8 *)x_1(D) + 3B], mov r1, r3 @, orr r0, r2, ip @ tmp171, tmp184, tmp194 ldmfd sp!, {r4, r5, r6, r7} bx lr In both cases the code is slightly suboptimal. One may wonder why wasting r2 with the constant 0 in the second case for example. And all the mov's could be folded in subsequent orr's, etc. Now with the asm-generic version: foo: ldr r0, [r0, #0] @ unaligned @,* x bx lr @ bar: mov r3, r0 @ x, x ldr r0, [r0, #0] @ unaligned @,* x ldr r1, [r3, aosp-mirror#4] @ unaligned @, bx lr @ This is way better of course, but only because this was compiled for ARMv7. In this case the compiler knows that the hardware can do unaligned word access. This isn't that obvious for foo(), but if we remove the get_unaligned() from bar as follows: long long bar (long long *x) {return *x; } then the resulting code is: bar: ldmia r0, {r0, r1} @ x,, bx lr @ So this proves that the presumed aligned vs unaligned cases does have influence on the instructions the compiler may use and that the above unaligned code results are not just an accident. Still... this isn't fully conclusive without at least looking at the resulting assembly fron a pre ARMv6 compilation. Let's see with an ARMv5 target: foo: ldrb r3, [r0, #0] @ zero_extendqisi2 @ tmp139,* x ldrb r1, [r0, stratosk#1] @ zero_extendqisi2 @ tmp140, ldrb r2, [r0, stratosk#2] @ zero_extendqisi2 @ tmp143, ldrb r0, [r0, stratosk#3] @ zero_extendqisi2 @ tmp146, orr r3, r3, r1, asl aosp-mirror#8 @, tmp142, tmp139, tmp140, orr r3, r3, r2, asl #16 @, tmp145, tmp142, tmp143, orr r0, r3, r0, asl #24 @,, tmp145, tmp146, bx lr @ bar: stmfd sp!, {r4, r5, r6, r7} @, ldrb r2, [r0, #0] @ zero_extendqisi2 @ tmp139,* x ldrb r7, [r0, stratosk#1] @ zero_extendqisi2 @ tmp140, ldrb r3, [r0, aosp-mirror#4] @ zero_extendqisi2 @ tmp149, ldrb r6, [r0, aosp-mirror#5] @ zero_extendqisi2 @ tmp150, ldrb r5, [r0, stratosk#2] @ zero_extendqisi2 @ tmp143, ldrb r4, [r0, aosp-mirror#6] @ zero_extendqisi2 @ tmp153, ldrb r1, [r0, aosp-mirror#7] @ zero_extendqisi2 @ tmp156, ldrb ip, [r0, stratosk#3] @ zero_extendqisi2 @ tmp146, orr r2, r2, r7, asl aosp-mirror#8 @, tmp142, tmp139, tmp140, orr r3, r3, r6, asl aosp-mirror#8 @, tmp152, tmp149, tmp150, orr r2, r2, r5, asl #16 @, tmp145, tmp142, tmp143, orr r3, r3, r4, asl #16 @, tmp155, tmp152, tmp153, orr r0, r2, ip, asl #24 @,, tmp145, tmp146, orr r1, r3, r1, asl #24 @,, tmp155, tmp156, ldmfd sp!, {r4, r5, r6, r7} bx lr Compared to the initial results, this is really nicely optimized and I couldn't do much better if I were to hand code it myself. Signed-off-by: Rob Herring <rob.herring@calxeda.com> Reviewed-by: Nicolas Pitre <nico@linaro.org> Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> modified for Mako from kernel.org reference Signed-off-by: faux123 <reioux@gmail.com>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 20, 2016
Previous commit was incomplete Signed-off-by: Park Ju Hyung <qkrwngud825@gmail.com>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 20, 2016
With the introduction of fair queued rwlock, recursive read_lock() may hang the offending process if there is a write_lock() somewhere in between. With recursive read_lock checking enabled, the following error was reported: ============================================= [ INFO: possible recursive locking detected ] 3.16.0-rc1 stratosk#2 Tainted: G E --------------------------------------------- load_policy/708 is trying to acquire lock: (policy_rwlock){.+.+..}, at: [<ffffffff8125b32a>] security_genfs_sid+0x3a/0x170 but task is already holding lock: (policy_rwlock){.+.+..}, at: [<ffffffff8125b48c>] security_fs_use+0x2c/0x110 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(policy_rwlock); lock(policy_rwlock); This patch fixes the occurrence of recursive read_lock() of policy_rwlock by adding a helper function __security_genfs_sid() which requires caller to take the lock before calling it. The security_fs_use() was then modified to call the new helper function. Signed-off-by: Waiman Long <Waiman.Long@hp.com> Acked-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Paul Moore <pmoore@redhat.com> Signed-off-by: Pranav Vashi <neobuddy89@gmail.com>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 20, 2016
msm_sat_enqueue() calls spin_lock() and msm_sat_dequeue() calls spin_lock_irqsave(). This leads to lockdep warnings about the same lock being taken in interrupts on and interrupts off context which can lead to a potential deadlock. ================================= [ INFO: inconsistent lock state ] 3.4.0+ #382 Tainted: G W --------------------------------- inconsistent {HARDIRQ-ON-W} -> {IN-HARDIRQ-W} usage. kworker/u:2/94 [HC1[1]:SC0[0]:HE0:SE1] takes: (&(&sat->lock)->rlock){?.+...}, at: [<c0348b80>] msm_sat_enqueue+0x20/0xb0 {HARDIRQ-ON-W} state was registered at: [<c00bd290>] __lock_acquire+0x664/0x8d8 [<c00bd690>] lock_acquire+0x18c/0x1e8 [<c06989a4>] _raw_spin_lock+0x38/0x48 [<c0348b80>] msm_sat_enqueue+0x20/0xb0 [<c0349338>] msm_slim_rxwq+0x278/0x42c [<c0349580>] msm_slim_rx_msgq_thread+0x94/0x1f8 [<c008e480>] kthread+0x90/0xa0 [<c000f438>] kernel_thread_exit+0x0/0x8 irq event stamp: 24219 hardirqs last enabled at (24218): [<c0699188>] _raw_spin_unlock_irqrestore+0x3c/0x68 hardirqs last disabled at (24219): [<c06993f4>] __irq_svc+0x34/0x78 softirqs last enabled at (24129): [<c00738a8>] irq_exit+0x54/0xa8 softirqs last disabled at (24108): [<c00738a8>] irq_exit+0x54/0xa8 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&(&sat->lock)->rlock); <Interrupt> lock(&(&sat->lock)->rlock); *** DEADLOCK *** 3 locks held by kworker/u:2/94: #0: ((sat->satcl.name)){.+.+.+}, at: [<c0087124>] process_one_work+0x1e8/0x648 stratosk#1: ((&sat->wd)){+.+.+.}, at: [<c0087124>] process_one_work+0x1e8/0x648 stratosk#2: (&dev->tx_lock){+.+...}, at: [<c0349cac>] msm_xfer_msg+0xb4/0x51c stack backtrace: [<c00151b0>] (unwind_backtrace+0x0/0x120) from [<c00b9d48>] (print_usage_bug+0x258/0x2c0) [<c00b9d48>] (print_usage_bug+0x258/0x2c0) from [<c00ba12c>] (mark_lock+0x37c/0x68c) [<c00ba12c>] (mark_lock+0x37c/0x68c) from [<c00bd20c>] (__lock_acquire+0x5e0/0x8d8) [<c00bd20c>] (__lock_acquire+0x5e0/0x8d8) from [<c00bd690>] (lock_acquire+0x18c/0x1e8) [<c00bd690>] (lock_acquire+0x18c/0x1e8) from [<c06989a4>] (_raw_spin_lock+0x38/0x48) [<c06989a4>] (_raw_spin_lock+0x38/0x48) from [<c0348b80>] (msm_sat_enqueue+0x20/0xb0) [<c0348b80>] (msm_sat_enqueue+0x20/0xb0) from [<c0348dc8>] (msm_slim_interrupt+0x1b8/0x4b0) [<c0348dc8>] (msm_slim_interrupt+0x1b8/0x4b0) from [<c00d2ccc>] (handle_irq_event_percpu+0x118/0x3a4) [<c00d2ccc>] (handle_irq_event_percpu+0x118/0x3a4) from [<c00d2f94>] (handle_irq_event+0x3c/0x5c) [<c00d2f94>] (handle_irq_event+0x3c/0x5c) from [<c00d5c4c>] (handle_fasteoi_irq+0xd0/0x11c) [<c00d5c4c>] (handle_fasteoi_irq+0xd0/0x11c) from [<c00d2ae4>] (generic_handle_irq+0x24/0x2c) [<c00d2ae4>] (generic_handle_irq+0x24/0x2c) from [<c000f370>] (handle_IRQ+0x7c/0xc0) [<c000f370>] (handle_IRQ+0x7c/0xc0) from [<c0008774>] (gic_handle_irq+0x6c/0xc4) [<c0008774>] (gic_handle_irq+0x6c/0xc4) from [<c0699404>] (__irq_svc+0x44/0x78) Exception stack(0xee061cd8 to 0xee061d20) 1cc0: 00000001 eebda7c8 1ce0: 00000000 eebda400 20000013 c12b5430 ec96c168 c12b542c c12b5430 00000001 1d00: 20000013 ee061e14 3eb13eb1 ee061d20 c00ba520 c069918c 20000013 ffffffff [<c0699404>] (__irq_svc+0x44/0x78) from [<c069918c>] (_raw_spin_unlock_irqrestore+0x40/0x68) [<c069918c>] (_raw_spin_unlock_irqrestore+0x40/0x68) from [<c0279f2c>] (__debug_object_init+0x30c/0x344) [<c0279f2c>] (__debug_object_init+0x30c/0x344) from [<c007b068>] (init_timer_on_stack_key+0x18/0x30) [<c007b068>] (init_timer_on_stack_key+0x18/0x30) from [<c0695248>] (schedule_timeout+0x8c/0x4c0) [<c0695248>] (schedule_timeout+0x8c/0x4c0) from [<c069768c>] (wait_for_common+0xec/0x164) [<c069768c>] (wait_for_common+0xec/0x164) from [<c0349fc4>] (msm_xfer_msg+0x3cc/0x51c) [<c0349fc4>] (msm_xfer_msg+0x3cc/0x51c) from [<c034a358>] (slim_sat_rxprocess+0x244/0x664) [<c034a358>] (slim_sat_rxprocess+0x244/0x664) from [<c0087290>] (process_one_work+0x354/0x648) [<c0087290>] (process_one_work+0x354/0x648) from [<c0089754>] (worker_thread+0x1a8/0x2a8) [<c0089754>] (worker_thread+0x1a8/0x2a8) from [<c008e480>] (kthread+0x90/0xa0) [<c008e480>] (kthread+0x90/0xa0) from [<c000f438>] (kernel_thread_exit+0x0/0x8) Make this lock irqsafe as well so that this potential bug doesn't occur. Change-Id: Icbef6d1d749ee6ee81b079e19e57f22c38f00c68 Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 20, 2016
workqueue: change BUG_ON() to WARN_ON() This BUG_ON() can be triggered if you call schedule_work() before calling INIT_WORK(). It is a bug definitely, but it's nicer to just print a stack trace and return. Reported-by: Matt Renzelmann <mjr@cs.wisc.edu> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: Catch more locking problems with flush_work() If a workqueue is flushed with flush_work() lockdep checking can be circumvented. For example: static DEFINE_MUTEX(mutex); static void my_work(struct work_struct *w) { mutex_lock(&mutex); mutex_unlock(&mutex); } static DECLARE_WORK(work, my_work); static int __init start_test_module(void) { schedule_work(&work); return 0; } module_init(start_test_module); static void __exit stop_test_module(void) { mutex_lock(&mutex); flush_work(&work); mutex_unlock(&mutex); } module_exit(stop_test_module); would not always print a warning when flush_work() was called. In this trivial example nothing could go wrong since we are guaranteed module_init() and module_exit() don't run concurrently, but if the work item is schedule asynchronously we could have a scenario where the work item is running just at the time flush_work() is called resulting in a classic ABBA locking problem. Add a lockdep hint by acquiring and releasing the work item lockdep_map in flush_work() so that we always catch this potential deadlock scenario. Signed-off-by: Stephen Boyd <sboyd@codeaurora.org> Reviewed-by: Yong Zhang <yong.zhang0@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> lockdep: fix oops in processing workqueue Under memory load, on x86_64, with lockdep enabled, the workqueue's process_one_work() has been seen to oops in __lock_acquire(), barfing on a 0xffffffff00000000 pointer in the lockdep_map's class_cache[]. Because it's permissible to free a work_struct from its callout function, the map used is an onstack copy of the map given in the work_struct: and that copy is made without any locking. Surprisingly, gcc (4.5.1 in Hugh's case) uses "rep movsl" rather than "rep movsq" for that structure copy: which might race with a workqueue user's wait_on_work() doing lock_map_acquire() on the source of the copy, putting a pointer into the class_cache[], but only in time for the top half of that pointer to be copied to the destination map. Boom when process_one_work() subsequently does lock_map_acquire() on its onstack copy of the lockdep_map. Fix this, and a similar instance in call_timer_fn(), with a lockdep_copy_map() function which additionally NULLs the class_cache[]. Note: this oops was actually seen on 3.4-next, where flush_work() newly does the racing lock_map_acquire(); but Tejun points out that 3.4 and earlier are already vulnerable to the same through wait_on_work(). * Patch orginally from Peter. Hugh modified it a bit and wrote the description. Signed-off-by: Peter Zijlstra <peterz@infradead.org> Reported-by: Hugh Dickins <hughd@google.com> LKML-Reference: <alpine.LSU.2.00.1205070951170.1544@eggly.anvils> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: perform cpu down operations from low priority cpu_notifier() Currently, all workqueue cpu hotplug operations run off CPU_PRI_WORKQUEUE which is higher than normal notifiers. This is to ensure that workqueue is up and running while bringing up a CPU before other notifiers try to use workqueue on the CPU. Per-cpu workqueues are supposed to remain working and bound to the CPU for normal CPU_DOWN_PREPARE notifiers. This holds mostly true even with workqueue offlining running with higher priority because workqueue CPU_DOWN_PREPARE only creates a bound trustee thread which runs the per-cpu workqueue without concurrency management without explicitly detaching the existing workers. However, if the trustee needs to create new workers, it creates unbound workers which may wander off to other CPUs while CPU_DOWN_PREPARE notifiers are in progress. Furthermore, if the CPU down is cancelled, the per-CPU workqueue may end up with workers which aren't bound to the CPU. While reliably reproducible with a convoluted artificial test-case involving scheduling and flushing CPU burning work items from CPU down notifiers, this isn't very likely to happen in the wild, and, even when it happens, the effects are likely to be hidden by the following successful CPU down. Fix it by using different priorities for up and down notifiers - high priority for up operations and low priority for down operations. Workqueue cpu hotplug operations will soon go through further cleanup. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: drop CPU_DYING notifier operation Workqueue used CPU_DYING notification to mark GCWQ_DISASSOCIATED. This was necessary because workqueue's CPU_DOWN_PREPARE happened before other DOWN_PREPARE notifiers and workqueue needed to stay associated across the rest of DOWN_PREPARE. After the previous patch, workqueue's DOWN_PREPARE happens after others and can set GCWQ_DISASSOCIATED directly. Drop CPU_DYING and let the trustee set GCWQ_DISASSOCIATED after disabling concurrency management. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: ROGUE workers are UNBOUND workers Currently, WORKER_UNBOUND is used to mark workers for the unbound global_cwq and WORKER_ROGUE is used to mark workers for disassociated per-cpu global_cwqs. Both are used to make the marked worker skip concurrency management and the only place they make any difference is in worker_enter_idle() where WORKER_ROGUE is used to skip scheduling idle timer, which can easily be replaced with trustee state testing. This patch replaces WORKER_ROGUE with WORKER_UNBOUND and drops WORKER_ROGUE. This is to prepare for removing trustee and handling disassociated global_cwqs as unbound. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: use mutex for global_cwq manager exclusion POOL_MANAGING_WORKERS is used to ensure that at most one worker takes the manager role at any given time on a given global_cwq. Trustee later hitched on it to assume manager adding blocking wait for the bit. As trustee already needed a custom wait mechanism, waiting for MANAGING_WORKERS was rolled into the same mechanism. Trustee is scheduled to be removed. This patch separates out MANAGING_WORKERS wait into per-pool mutex. Workers use mutex_trylock() to test for manager role and trustee uses mutex_lock() to claim manager roles. gcwq_claim/release_management() helpers are added to grab and release manager roles of all pools on a global_cwq. gcwq_claim_management() always grabs pool manager mutexes in ascending pool index order and uses pool index as lockdep subclass. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: drop @bind from create_worker() Currently, create_worker()'s callers are responsible for deciding whether the newly created worker should be bound to the associated CPU and create_worker() sets WORKER_UNBOUND only for the workers for the unbound global_cwq. Creation during normal operation is always via maybe_create_worker() and @bind is true. For workers created during hotplug, @bind is false. Normal operation path is planned to be used even while the CPU is going through hotplug operations or offline and this static decision won't work. Drop @bind from create_worker() and decide whether to bind by looking at GCWQ_DISASSOCIATED. create_worker() will also set WORKER_UNBOUND autmatically if disassociated. To avoid flipping GCWQ_DISASSOCIATED while create_worker() is in progress, the flag is now allowed to be changed only while holding all manager_mutexes on the global_cwq. This requires that GCWQ_DISASSOCIATED is not cleared behind trustee's back. CPU_ONLINE no longer clears DISASSOCIATED before flushing trustee, which clears DISASSOCIATED before rebinding remaining workers if asked to release. For cases where trustee isn't around, CPU_ONLINE clears DISASSOCIATED after flushing trustee. Also, now, first_idle has UNBOUND set on creation which is explicitly cleared by CPU_ONLINE while binding it. These convolutions will soon be removed by further simplification of CPU hotplug path. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: reimplement CPU online rebinding to handle idle workers Currently, if there are left workers when a CPU is being brough back online, the trustee kills all idle workers and scheduled rebind_work so that they re-bind to the CPU after the currently executing work is finished. This works for busy workers because concurrency management doesn't try to wake up them from scheduler callbacks, which require the target task to be on the local run queue. The busy worker bumps concurrency counter appropriately as it clears WORKER_UNBOUND from the rebind work item and it's bound to the CPU before returning to the idle state. To reduce CPU on/offlining overhead (as many embedded systems use it for powersaving) and simplify the code path, workqueue is planned to be modified to retain idle workers across CPU on/offlining. This patch reimplements CPU online rebinding such that it can also handle idle workers. As noted earlier, due to the local wakeup requirement, rebinding idle workers is tricky. All idle workers must be re-bound before scheduler callbacks are enabled. This is achieved by interlocking idle re-binding. Idle workers are requested to re-bind and then hold until all idle re-binding is complete so that no bound worker starts executing work item. Only after all idle workers are re-bound and parked, CPU_ONLINE proceeds to release them and queue rebind work item to busy workers thus guaranteeing scheduler callbacks aren't invoked until all idle workers are ready. worker_rebind_fn() is renamed to busy_worker_rebind_fn() and idle_worker_rebind() for idle workers is added. Rebinding logic is moved to rebind_workers() and now called from CPU_ONLINE after flushing trustee. While at it, add CPU sanity check in worker_thread(). Note that now a worker may become idle or the manager between trustee release and rebinding during CPU_ONLINE. As the previous patch updated create_worker() so that it can be used by regular manager while unbound and this patch implements idle re-binding, this is safe. This prepares for removal of trustee and keeping idle workers across CPU hotplugs. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: don't butcher idle workers on an offline CPU Currently, during CPU offlining, after all pending work items are drained, the trustee butchers all workers. Also, on CPU onlining failure, workqueue_cpu_callback() ensures that the first idle worker is destroyed. Combined, these guarantee that an offline CPU doesn't have any worker for it once all the lingering work items are finished. This guarantee isn't really necessary and makes CPU on/offlining more expensive than needs to be, especially for platforms which use CPU hotplug for powersaving. This patch lets offline CPUs removes idle worker butchering from the trustee and let a CPU which failed onlining keep the created first worker. The first worker is created if the CPU doesn't have any during CPU_DOWN_PREPARE and started right away. If onlining succeeds, the rebind_workers() call in CPU_ONLINE will rebind it like any other workers. If onlining fails, the worker is left alone till the next try. This makes CPU hotplugs cheaper by allowing global_cwqs to keep workers across them and simplifies code. Note that trustee doesn't re-arm idle timer when it's done and thus the disassociated global_cwq will keep all workers until it comes back online. This will be improved by further patches. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: remove CPU offline trustee With the previous changes, a disassociated global_cwq now can run as an unbound one on its own - it can create workers as necessary to drain remaining works after the CPU has been brought down and manage the number of workers using the usual idle timer mechanism making trustee completely redundant except for the actual unbinding operation. This patch removes the trustee and let a disassociated global_cwq manage itself. Unbinding is moved to a work item (for CPU affinity) which is scheduled and flushed from CPU_DONW_PREPARE. This patch moves nr_running clearing outside gcwq and manager locks to simplify the code. As nr_running is unused at the point, this is safe. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: simplify CPU hotplug code With trustee gone, CPU hotplug code can be simplified. * gcwq_claim/release_management() now grab and release gcwq lock too respectively and gained _and_lock and _and_unlock postfixes. * All CPU hotplug logic was implemented in workqueue_cpu_callback() which was called by workqueue_cpu_up/down_callback() for the correct priority. This was because up and down paths shared a lot of logic, which is no longer true. Remove workqueue_cpu_callback() and move all hotplug logic into the two actual callbacks. This patch doesn't make any functional changes. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: fix spurious CPU locality WARN from process_one_work() 25511a4776 "workqueue: reimplement CPU online rebinding to handle idle workers" added CPU locality sanity check in process_one_work(). It triggers if a worker is executing on a different CPU without UNBOUND or REBIND set. This works for all normal workers but rescuers can trigger this spuriously when they're serving the unbound or a disassociated global_cwq - rescuers don't have either flag set and thus its gcwq->cpu can be a different value including %WORK_CPU_UNBOUND. Fix it by additionally testing %GCWQ_DISASSOCIATED. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> LKML-Refence: <20120721213656.GA7783@linux.vnet.ibm.com> workqueue: reorder queueing functions so that _on() variants are on top Currently, queue/schedule[_delayed]_work_on() are located below the counterpart without the _on postifx even though the latter is usually implemented using the former. Swap them. This is cleanup and doesn't cause any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: make queueing functions return bool All queueing functions return 1 on success, 0 if the work item was already pending. Update them to return bool instead. This signifies better that they don't return 0 / -errno. This is cleanup and doesn't cause any functional difference. While at it, fix comment opening for schedule_work_on(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: add missing smp_wmb() in process_one_work() WORK_STRUCT_PENDING is used to claim ownership of a work item and process_one_work() releases it before starting execution. When someone else grabs PENDING, all pre-release updates to the work item should be visible and all updates made by the new owner should happen afterwards. Grabbing PENDING uses test_and_set_bit() and thus has a full barrier; however, clearing doesn't have a matching wmb. Given the preceding spin_unlock and use of clear_bit, I don't believe this can be a problem on an actual machine and there hasn't been any related report but it still is theretically possible for clear_pending to permeate upwards and happen before work->entry update. Add an explicit smp_wmb() before work_clear_pending(). Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: stable@vger.kernel.org workqueue: disable irq while manipulating PENDING Queueing operations use WORK_STRUCT_PENDING_BIT to synchronize access to the target work item. They first try to claim the bit and proceed with queueing only after that succeeds and there's a window between PENDING being set and the actual queueing where the task can be interrupted or preempted. There's also a similar window in process_one_work() when clearing PENDING. A work item is dequeued, gcwq->lock is released and then PENDING is cleared and the worker might get interrupted or preempted between releasing gcwq->lock and clearing PENDING. cancel[_delayed]_work_sync() tries to claim or steal PENDING. The function assumes that a work item with PENDING is either queued or in the process of being [de]queued. In the latter case, it busy-loops until either the work item loses PENDING or is queued. If canceling coincides with the above described interrupts or preemptions, the canceling task will busy-loop while the queueing or executing task is preempted. This patch keeps irq disabled across claiming PENDING and actual queueing and moves PENDING clearing in process_one_work() inside gcwq->lock so that busy looping from PENDING && !queued doesn't wait for interrupted/preempted tasks. Note that, in process_one_work(), setting last CPU and clearing PENDING got merged into single operation. This removes possible long busy-loops and will allow using try_to_grab_pending() from bh and irq contexts. v2: __queue_work() was testing preempt_count() to ensure that the caller has disabled preemption. This triggers spuriously if !CONFIG_PREEMPT_COUNT. Use preemptible() instead. Reported by Fengguang Wu. v3: Disable irq instead of preemption. IRQ will be disabled while grabbing gcwq->lock later anyway and this allows using try_to_grab_pending() from bh and irq contexts. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Fengguang Wu <fengguang.wu@intel.com> workqueue: set delayed_work->timer function on initialization delayed_work->timer.function is currently initialized during queue_delayed_work_on(). Export delayed_work_timer_fn() and set delayed_work timer function during delayed_work initialization together with other fields. This ensures the timer function is always valid on an initialized delayed_work. This is to help mod_delayed_work() implementation. To detect delayed_work users which diddle with the internal timer, trigger WARN if timer function doesn't match on queue. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: unify local CPU queueing handling Queueing functions have been using different methods to determine the local CPU. * queue_work() superflously uses get/put_cpu() to acquire and hold the local CPU across queue_work_on(). * delayed_work_timer_fn() uses smp_processor_id(). * queue_delayed_work() calls queue_delayed_work_on() with -1 @cpu which is interpreted as the local CPU. * flush_delayed_work[_sync]() were using raw_smp_processor_id(). * __queue_work() interprets %WORK_CPU_UNBOUND as local CPU if the target workqueue is bound one but nobody uses this. This patch converts all functions to uniformly use %WORK_CPU_UNBOUND to indicate local CPU and use the local binding feature of __queue_work(). unlikely() is dropped from %WORK_CPU_UNBOUND handling in __queue_work(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix zero @delay handling of queue_delayed_work_on() If @delay is zero and the dealyed_work is idle, queue_delayed_work() queues it for immediate execution; however, queue_delayed_work_on() lacks this logic and always goes through timer regardless of @delay. This patch moves 0 @delay handling logic from queue_delayed_work() to queue_delayed_work_on() so that both functions behave the same. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: move try_to_grab_pending() upwards try_to_grab_pending() will be used by to-be-implemented mod_delayed_work[_on](). Move try_to_grab_pending() and related functions above queueing functions. This patch only moves functions around. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: introduce WORK_OFFQ_FLAG_* Low WORK_STRUCT_FLAG_BITS bits of work_struct->data contain WORK_STRUCT_FLAG_* and flush color. If the work item is queued, the rest point to the cpu_workqueue with WORK_STRUCT_CWQ set; otherwise, WORK_STRUCT_CWQ is clear and the bits contain the last CPU number - either a real CPU number or one of WORK_CPU_*. Scheduled addition of mod_delayed_work[_on]() requires an additional flag, which is used only while a work item is off queue. There are more than enough bits to represent off-queue CPU number on both 32 and 64bits. This patch introduces WORK_OFFQ_FLAG_* which occupy the lower part of the @work->data high bits while off queue. This patch doesn't define any actual OFFQ flag yet. Off-queue CPU number is now shifted by WORK_OFFQ_CPU_SHIFT, which adds the number of bits used by OFFQ flags to WORK_STRUCT_FLAG_SHIFT, to make room for OFFQ flags. To avoid shift width warning with large WORK_OFFQ_FLAG_BITS, ulong cast is added to WORK_STRUCT_NO_CPU and, just in case, BUILD_BUG_ON() to check that there are enough bits to accomodate off-queue CPU number is added. This patch doesn't make any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: factor out __queue_delayed_work() from queue_delayed_work_on() This is to prepare for mod_delayed_work[_on]() and doesn't cause any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reorganize try_to_grab_pending() and __cancel_timer_work() * Use bool @is_dwork instead of @timer and let try_to_grab_pending() use to_delayed_work() to determine the delayed_work address. * Move timer handling from __cancel_work_timer() to try_to_grab_pending(). * Make try_to_grab_pending() use -EAGAIN instead of -1 for busy-looping and drop the ret local variable. * Add proper function comment to try_to_grab_pending(). This makes the code a bit easier to understand and will ease further changes. This patch doesn't make any functional change. v2: Use @is_dwork instead of @timer. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: mark a work item being canceled as such There can be two reasons try_to_grab_pending() can fail with -EAGAIN. One is when someone else is queueing or deqeueing the work item. With the previous patches, it is guaranteed that PENDING and queued state will soon agree making it safe to busy-retry in this case. The other is if multiple __cancel_work_timer() invocations are racing one another. __cancel_work_timer() grabs PENDING and then waits for running instances of the target work item on all CPUs while holding PENDING and !queued. try_to_grab_pending() invoked from another task will keep returning -EAGAIN while the current owner is waiting. Not distinguishing the two cases is okay because __cancel_work_timer() is the only user of try_to_grab_pending() and it invokes wait_on_work() whenever grabbing fails. For the first case, busy looping should be fine but wait_on_work() doesn't cause any critical problem. For the latter case, the new contender usually waits for the same condition as the current owner, so no unnecessarily extended busy-looping happens. Combined, these make __cancel_work_timer() technically correct even without irq protection while grabbing PENDING or distinguishing the two different cases. While the current code is technically correct, not distinguishing the two cases makes it difficult to use try_to_grab_pending() for other purposes than canceling because it's impossible to tell whether it's safe to busy-retry grabbing. This patch adds a mechanism to mark a work item being canceled. try_to_grab_pending() now disables irq on success and returns -EAGAIN to indicate that grabbing failed but PENDING and queued states are gonna agree soon and it's safe to busy-loop. It returns -ENOENT if the work item is being canceled and it may stay PENDING && !queued for arbitrary amount of time. __cancel_work_timer() is modified to mark the work canceling with WORK_OFFQ_CANCELING after grabbing PENDING, thus making try_to_grab_pending() fail with -ENOENT instead of -EAGAIN. Also, it invokes wait_on_work() iff grabbing failed with -ENOENT. This isn't necessary for correctness but makes it consistent with other future users of try_to_grab_pending(). v2: try_to_grab_pending() was testing preempt_count() to ensure that the caller has disabled preemption. This triggers spuriously if !CONFIG_PREEMPT_COUNT. Use preemptible() instead. Reported by Fengguang Wu. v3: Updated so that try_to_grab_pending() disables irq on success rather than requiring preemption disabled by the caller. This makes busy-looping easier and will allow try_to_grap_pending() to be used from bh/irq contexts. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Fengguang Wu <fengguang.wu@intel.com> workqueue: implement mod_delayed_work[_on]() Workqueue was lacking a mechanism to modify the timeout of an already pending delayed_work. delayed_work users have been working around this using several methods - using an explicit timer + work item, messing directly with delayed_work->timer, and canceling before re-queueing, all of which are error-prone and/or ugly. This patch implements mod_delayed_work[_on]() which behaves similarly to mod_timer() - if the delayed_work is idle, it's queued with the given delay; otherwise, its timeout is modified to the new value. Zero @delay guarantees immediate execution. v2: Updated to reflect try_to_grab_pending() changes. Now safe to be called from bh context. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> workqueue: fix CPU binding of flush_delayed_work[_sync]() delayed_work encodes the workqueue to use and the last CPU in delayed_work->work.data while it's on timer. The target CPU is implicitly recorded as the CPU the timer is queued on and delayed_work_timer_fn() queues delayed_work->work to the CPU it is running on. Unfortunately, this leaves flush_delayed_work[_sync]() no way to find out which CPU the delayed_work was queued for when they try to re-queue after killing the timer. Currently, it chooses the local CPU flush is running on. This can unexpectedly move a delayed_work queued on a specific CPU to another CPU and lead to subtle errors. There isn't much point in trying to save several bytes in struct delayed_work, which is already close to a hundred bytes on 64bit with all debug options turned off. This patch adds delayed_work->cpu to remember the CPU it's queued for. Note that if the timer is migrated during CPU down, the work item could be queued to the downed global_cwq after this change. As a detached global_cwq behaves like an unbound one, this doesn't change much for the delayed_work. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> workqueue: add missing wmb() in clear_work_data() Any operation which clears PENDING should be preceded by a wmb to guarantee that the next PENDING owner sees all the changes made before PENDING release. There are only two places where PENDING is cleared - set_work_cpu_and_clear_pending() and clear_work_data(). The caller of the former already does smp_wmb() but the latter doesn't have any. Move the wmb above set_work_cpu_and_clear_pending() into it and add one to clear_work_data(). There hasn't been any report related to this issue, and, given how clear_work_data() is used, it is extremely unlikely to have caused any actual problems on any architecture. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> workqueue: use enum value to set array size of pools in gcwq Commit 3270476a6c0ce322354df8679652f060d66526dc ('workqueue: reimplement WQ_HIGHPRI using a separate worker_pool') introduce separate worker_pool for HIGHPRI. Although there is NR_WORKER_POOLS enum value which represent size of pools, definition of worker_pool in gcwq doesn't use it. Using it makes code robust and prevent future mistakes. So change code to use this enum value. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: correct req_cpu in trace_workqueue_queue_work() When we do tracing workqueue_queue_work(), it records requested cpu. But, if !(@wq->flag & WQ_UNBOUND) and @cpu is WORK_CPU_UNBOUND, requested cpu is changed as local cpu. In case of @wq->flag & WQ_UNBOUND, above change is not occured, therefore it is reasonable to correct it. Use temporary local variable for storing requested cpu. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: change value of lcpu in __queue_delayed_work_on() We assign cpu id into work struct's data field in __queue_delayed_work_on(). In current implementation, when work is come in first time, current running cpu id is assigned. If we do __queue_delayed_work_on() with CPU A on CPU B, __queue_work() invoked in delayed_work_timer_fn() go into the following sub-optimal path in case of WQ_NON_REENTRANT. gcwq = get_gcwq(cpu); if (wq->flags & WQ_NON_REENTRANT && (last_gcwq = get_work_gcwq(work)) && last_gcwq != gcwq) { Change lcpu to @cpu and rechange lcpu to local cpu if lcpu is WORK_CPU_UNBOUND. It is sufficient to prevent to go into sub-optimal path. tj: Slightly rephrased the comment. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: introduce system_highpri_wq Commit 3270476a6c0ce322354df8679652f060d66526dc ('workqueue: reimplement WQ_HIGHPRI using a separate worker_pool') introduce separate worker pool for HIGHPRI. When we handle busyworkers for gcwq, it can be normal worker or highpri worker. But, we don't consider this difference in rebind_workers(), we use just system_wq for highpri worker. It makes mismatch between cwq->pool and worker->pool. It doesn't make error in current implementation, but possible in the future. Now, we introduce system_highpri_wq to use proper cwq for highpri workers in rebind_workers(). Following patch fix this issue properly. tj: Even apart from rebinding, having system_highpri_wq generally makes sense. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use system_highpri_wq for highpri workers in rebind_workers() In rebind_workers(), we do inserting a work to rebind to cpu for busy workers. Currently, in this case, we use only system_wq. This makes a possible error situation as there is mismatch between cwq->pool and worker->pool. To prevent this, we should use system_highpri_wq for highpri worker to match theses. This implements it. tj: Rephrased comment a bit. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use system_highpri_wq for unbind_work To speed cpu down processing up, use system_highpri_wq. As scheduling priority of workers on it is higher than system_wq and it is not contended by other normal works on this cpu, work on it is processed faster than system_wq. tj: CPU up/downs care quite a bit about latency these days. This shouldn't hurt anything and makes sense. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix checkpatch issues Fixed some checkpatch warnings. tj: adapted to wq/for-3.7 and massaged pr_xxx() format strings a bit. Signed-off-by: Valentin Ilie <valentin.ilie@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> LKML-Reference: <1345326762-21747-1-git-send-email-valentin.ilie@gmail.com> workqueue: make all workqueues non-reentrant By default, each per-cpu part of a bound workqueue operates separately and a work item may be executing concurrently on different CPUs. The behavior avoids some cross-cpu traffic but leads to subtle weirdities and not-so-subtle contortions in the API. * There's no sane usefulness in allowing a single work item to be executed concurrently on multiple CPUs. People just get the behavior unintentionally and get surprised after learning about it. Most either explicitly synchronize or use non-reentrant/ordered workqueue but this is error-prone. * flush_work() can't wait for multiple instances of the same work item on different CPUs. If a work item is executing on cpu0 and then queued on cpu1, flush_work() can only wait for the one on cpu1. Unfortunately, work items can easily cross CPU boundaries unintentionally when the queueing thread gets migrated. This means that if multiple queuers compete, flush_work() can't even guarantee that the instance queued right before it is finished before returning. * flush_work_sync() was added to work around some of the deficiencies of flush_work(). In addition to the usual flushing, it ensures that all currently executing instances are finished before returning. This operation is expensive as it has to walk all CPUs and at the same time fails to address competing queuer case. Incorrectly using flush_work() when flush_work_sync() is necessary is an easy error to make and can lead to bugs which are difficult to reproduce. * Similar problems exist for flush_delayed_work[_sync](). Other than the cross-cpu access concern, there's no benefit in allowing parallel execution and it's plain silly to have this level of contortion for workqueue which is widely used from core code to extremely obscure drivers. This patch makes all workqueues non-reentrant. If a work item is executing on a different CPU when queueing is requested, it is always queued to that CPU. This guarantees that any given work item can be executing on one CPU at maximum and if a work item is queued and executing, both are on the same CPU. The only behavior change which may affect workqueue users negatively is that non-reentrancy overrides the affinity specified by queue_work_on(). On a reentrant workqueue, the affinity specified by queue_work_on() is always followed. Now, if the work item is executing on one of the CPUs, the work item will be queued there regardless of the requested affinity. I've reviewed all workqueue users which request explicit affinity, and, fortunately, none seems to be crazy enough to exploit parallel execution of the same work item. This adds an additional busy_hash lookup if the work item was previously queued on a different CPU. This shouldn't be noticeable under any sane workload. Work item queueing isn't a very high-frequency operation and they don't jump across CPUs all the time. In a micro benchmark to exaggerate this difference - measuring the time it takes for two work items to repeatedly jump between two CPUs a number (10M) of times with busy_hash table densely populated, the difference was around 3%. While the overhead is measureable, it is only visible in pathological cases and the difference isn't huge. This change brings much needed sanity to workqueue and makes its behavior consistent with timer. I think this is the right tradeoff to make. This enables significant simplification of workqueue API. Simplification patches will follow. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: gut flush[_delayed]_work_sync() Now that all workqueues are non-reentrant, flush[_delayed]_work_sync() are equivalent to flush[_delayed]_work(). Drop the separate implementation and make them thin wrappers around flush[_delayed]_work(). * start_flush_work() no longer takes @wait_executing as the only left user - flush_work() - always sets it to %true. * __cancel_work_timer() uses flush_work() instead of wait_on_work(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: gut system_nrt[_freezable]_wq() Now that all workqueues are non-reentrant, system[_freezable]_wq() are equivalent to system_nrt[_freezable]_wq(). Replace the latter with wrappers around system[_freezable]_wq(). The wrapping goes through inline functions so that __deprecated can be added easily. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: cosmetic whitespace updates for macro definitions Consistently use the last tab position for '\' line continuation in complex macro definitions. This is to help the following patches. This patch is cosmetic. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use hotcpu_notifier() for workqueue_cpu_down_callback() workqueue_cpu_down_callback() is used only if HOTPLUG_CPU=y, so hotcpu_notifier() fits better than cpu_notifier(). When HOTPLUG_CPU=y, hotcpu_notifier() and cpu_notifier() are the same. When HOTPLUG_CPU=n, if we use cpu_notifier(), workqueue_cpu_down_callback() will be called during boot to do nothing, and the memory of workqueue_cpu_down_callback() and gcwq_unbind_fn() will be discarded after boot. If we use hotcpu_notifier(), we can avoid the no-op call of workqueue_cpu_down_callback() and the memory of workqueue_cpu_down_callback() and gcwq_unbind_fn() will be discard at build time: $ ls -l kernel/workqueue.o.cpu_notifier kernel/workqueue.o.hotcpu_notifier -rw-rw-r-- 1 laijs laijs 484080 Sep 15 11:31 kernel/workqueue.o.cpu_notifier -rw-rw-r-- 1 laijs laijs 478240 Sep 15 11:31 kernel/workqueue.o.hotcpu_notifier $ size kernel/workqueue.o.cpu_notifier kernel/workqueue.o.hotcpu_notifier text data bss dec hex filename 18513 2387 1221 22121 5669 kernel/workqueue.o.cpu_notifier 18082 2355 1221 21658 549a kernel/workqueue.o.hotcpu_notifier tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reimplement cancel_delayed_work() using try_to_grab_pending() cancel_delayed_work() can't be called from IRQ handlers due to its use of del_timer_sync() and can't cancel work items which are already transferred from timer to worklist. Also, unlike other flush and cancel functions, a canceled delayed_work would still point to the last associated cpu_workqueue. If the workqueue is destroyed afterwards and the work item is re-used on a different workqueue, the queueing code can oops trying to dereference already freed cpu_workqueue. This patch reimplements cancel_delayed_work() using try_to_grab_pending() and set_work_cpu_and_clear_pending(). This allows the function to be called from IRQ handlers and makes its behavior consistent with other flush / cancel functions. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> workqueue: UNBOUND -> REBIND morphing in rebind_workers() should be atomic The compiler may compile the following code into TWO write/modify instructions. worker->flags &= ~WORKER_UNBOUND; worker->flags |= WORKER_REBIND; so the other CPU may temporarily see worker->flags which doesn't have either WORKER_UNBOUND or WORKER_REBIND set and perform local wakeup prematurely. Fix it by using single explicit assignment via ACCESS_ONCE(). Because idle workers have another WORKER_NOT_RUNNING flag, this bug doesn't exist for them; however, update it to use the same pattern for consistency. tj: Applied the change to idle workers too and updated comments and patch description a bit. Change-Id: I9b95f51d146c40c31ba028668d6f412bd74c6026 Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org workqueue: move WORKER_REBIND clearing in rebind_workers() to the end of the function This doesn't make any functional difference and is purely to help the next patch to be simpler. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> workqueue: fix possible deadlock in idle worker rebinding Currently, rebind_workers() and idle_worker_rebind() are two-way interlocked. rebind_workers() waits for idle workers to finish rebinding and rebound idle workers wait for rebind_workers() to finish rebinding busy workers before proceeding. Unfortunately, this isn't enough. The second wait from idle workers is implemented as follows. wait_event(gcwq->rebind_hold, !(worker->flags & WORKER_REBIND)); rebind_workers() clears WORKER_REBIND, wakes up the idle workers and then returns. If CPU hotplug cycle happens again before one of the idle workers finishes the above wait_event(), rebind_workers() will repeat the first part of the handshake - set WORKER_REBIND again and wait for the idle worker to finish rebinding - and this leads to deadlock because the idle worker would be waiting for WORKER_REBIND to clear. This is fixed by adding another interlocking step at the end - rebind_workers() now waits for all the idle workers to finish the above WORKER_REBIND wait before returning. This ensures that all rebinding steps are complete on all idle workers before the next hotplug cycle can happen. This problem was diagnosed by Lai Jiangshan who also posted a patch to fix the issue, upon which this patch is based. This is the minimal fix and further patches are scheduled for the next merge window to simplify the CPU hotplug path. Signed-off-by: Tejun Heo <tj@kernel.org> Original-patch-by: Lai Jiangshan <laijs@cn.fujitsu.com> LKML-Reference: <1346516916-1991-3-git-send-email-laijs@cn.fujitsu.com> workqueue: restore POOL_MANAGING_WORKERS This patch restores POOL_MANAGING_WORKERS which was replaced by pool->manager_mutex by 6037315269 "workqueue: use mutex for global_cwq manager exclusion". There's a subtle idle worker depletion bug across CPU hotplug events and we need to distinguish an actual manager and CPU hotplug preventing management. POOL_MANAGING_WORKERS will be used for the former and manager_mutex the later. This patch just lays POOL_MANAGING_WORKERS on top of the existing manager_mutex and doesn't introduce any synchronization changes. The next patch will update it. Note that this patch fixes a non-critical anomaly where too_many_workers() may return %true spuriously while CPU hotplug is in progress. While the issue could schedule idle timer spuriously, it didn't trigger any actual misbehavior. tj: Rewrote patch description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix possible idle worker depletion across CPU hotplug To simplify both normal and CPU hotplug paths, worker management is prevented while CPU hoplug is in progress. This is achieved by CPU hotplug holding the same exclusion mechanism used by workers to ensure there's only one manager per pool. If someone else seems to be performing the manager role, workers proceed to execute work items. CPU hotplug using the same mechanism can lead to idle worker depletion because all workers could proceed to execute work items while CPU hotplug is in progress and CPU hotplug itself wouldn't actually perform the worker management duty - it doesn't guarantee that there's an idle worker left when it releases management. This idle worker depletion, under extreme circumstances, can break forward-progress guarantee and thus lead to deadlock. This patch fixes the bug by using separate mechanisms for manager exclusion among workers and hotplug exclusion. For manager exclusion, POOL_MANAGING_WORKERS which was restored by the previous patch is used. pool->manager_mutex is now only used for exclusion between the elected manager and CPU hotplug. The elected manager won't proceed without holding pool->manager_mutex. This ensures that the worker which won the manager position can't skip managing while CPU hotplug is in progress. It will block on manager_mutex and perform management after CPU hotplug is complete. Note that hotplug may happen while waiting for manager_mutex. A manager isn't either on idle or busy list and thus the hoplug code can't unbind/rebind it. Make the manager handle its own un/rebinding. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: always clear WORKER_REBIND in busy_worker_rebind_fn() busy_worker_rebind_fn() didn't clear WORKER_REBIND if rebinding failed (CPU is down again). This used to be okay because the flag wasn't used for anything else. However, after 25511a477 "workqueue: reimplement CPU online rebinding to handle idle workers", WORKER_REBIND is also used to command idle workers to rebind. If not cleared, the worker may confuse the next CPU_UP cycle by having REBIND spuriously set or oops / get stuck by prematurely calling idle_worker_rebind(). WARNING: at /work/os/wq/kernel/workqueue.c:1323 worker_thread+0x4cd/0x5 00() Hardware name: Bochs Modules linked in: test_wq(O-) Pid: 33, comm: kworker/1:1 Tainted: G O 3.6.0-rc1-work+ #3 Call Trace: [<ffffffff8109039f>] warn_slowpath_common+0x7f/0xc0 [<ffffffff810903fa>] warn_slowpath_null+0x1a/0x20 [<ffffffff810b3f1d>] worker_thread+0x4cd/0x500 [<ffffffff810bc16e>] kthread+0xbe/0xd0 [<ffffffff81bd2664>] kernel_thread_helper+0x4/0x10 ---[ end trace e977cf20f4661968 ]--- BUG: unable to handle kernel NULL pointer dereference at (null) IP: [<ffffffff810b3db0>] worker_thread+0x360/0x500 PGD 0 Oops: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC Modules linked in: test_wq(O-) CPU 0 Pid: 33, comm: kworker/1:1 Tainted: G W O 3.6.0-rc1-work+ #3 Bochs Bochs RIP: 0010:[<ffffffff810b3db0>] [<ffffffff810b3db0>] worker_thread+0x360/0x500 RSP: 0018:ffff88001e1c9de0 EFLAGS: 00010086 RAX: 0000000000000000 RBX: ffff88001e633e00 RCX: 0000000000004140 RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000009 RBP: ffff88001e1c9ea0 R08: 0000000000000000 R09: 0000000000000001 R10: 0000000000000002 R11: 0000000000000000 R12: ffff88001fc8d580 R13: ffff88001fc8d590 R14: ffff88001e633e20 R15: ffff88001e1c6900 FS: 0000000000000000(0000) GS:ffff88001fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 0000000000000000 CR3: 00000000130e8000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process kworker/1:1 (pid: 33, threadinfo ffff88001e1c8000, task ffff88001e1c6900) Stack: ffff880000000000 ffff88001e1c9e40 0000000000000001 ffff88001e1c8010 ffff88001e519c78 ffff88001e1c9e58 ffff88001e1c6900 ffff88001e1c6900 ffff88001e1c6900 ffff88001e1c6900 ffff88001fc8d340 ffff88001fc8d340 Call Trace: [<ffffffff810bc16e>] kthread+0xbe/0xd0 [<ffffffff81bd2664>] kernel_thread_helper+0x4/0x10 Code: b1 00 f6 43 48 02 0f 85 91 01 00 00 48 8b 43 38 48 89 df 48 8b 00 48 89 45 90 e8 ac f0 ff ff 3c 01 0f 85 60 01 00 00 48 8b 53 50 <8b> 02 83 e8 01 85 c0 89 02 0f 84 3b 01 00 00 48 8b 43 38 48 8b RIP [<ffffffff810b3db0>] worker_thread+0x360/0x500 RSP <ffff88001e1c9de0> CR2: 0000000000000000 There was no reason to keep WORKER_REBIND on failure in the first place - WORKER_UNBOUND is guaranteed to be set in such cases preventing incorrectly activating concurrency management. Always clear WORKER_REBIND. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reimplement idle worker rebinding Currently rebind_workers() uses rebinds idle workers synchronously before proceeding to requesting busy workers to rebind. This is necessary because all workers on @worker_pool->idle_list must be bound before concurrency management local wake-ups from the busy workers take place. Unfortunately, the synchronous idle rebinding is quite complicated. This patch reimplements idle rebinding to simplify the code path. Rather than trying to make all idle workers bound before rebinding busy workers, we simply remove all to-be-bound idle workers from the idle list and let them add themselves back after completing rebinding (successful or not). As only workers which finished rebinding can on on the idle worker list, the idle worker list is guaranteed to have only bound workers unless CPU went down again and local wake-ups are safe. After the change, @worker_pool->nr_idle may deviate than the actual number of idle workers on @worker_pool->idle_list. More specifically, nr_idle may be non-zero while ->idle_list is empty. All users of ->nr_idle and ->idle_list are audited. The only affected one is too_many_workers() which is updated to check %false if ->idle_list is empty regardless of ->nr_idle. After this patch, rebind_workers() no longer performs the nasty idle-rebind retries which require temporary release of gcwq->lock, and both unbinding and rebinding are atomic w.r.t. global_cwq->lock. worker->idle_rebind and global_cwq->rebind_hold are now unnecessary and removed along with the definition of struct idle_rebind. Changed from V1: 1) remove unlikely from too_many_workers(), ->idle_list can be empty anytime, even before this patch, no reason to use unlikely. 2) fix a small rebasing mistake. (which is from rebasing the orignal fixing patch to for-next) 3) add a lot of comments. 4) clear WORKER_REBIND unconditionaly in idle_worker_rebind() tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: WORKER_REBIND is no longer necessary for busy rebinding Because the old unbind/rebinding implementation wasn't atomic w.r.t. GCWQ_DISASSOCIATED manipulation which is protected by global_cwq->lock, we had to use two flags, WORKER_UNBOUND and WORKER_REBIND, to avoid incorrectly losing all NOT_RUNNING bits with back-to-back CPU hotplug operations; otherwise, completion of rebinding while another unbinding is in progress could clear UNBIND prematurely. Now that both unbind/rebinding are atomic w.r.t. GCWQ_DISASSOCIATED, there's no need to use two flags. Just one is enough. Don't use WORKER_REBIND for busy rebinding. tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: WORKER_REBIND is no longer necessary for idle rebinding Now both worker destruction and idle rebinding remove the worker from idle list while it's still idle, so list_empty(&worker->entry) can be used to test whether either is pending and WORKER_DIE to distinguish between the two instead making WORKER_REBIND unnecessary. Use list_empty(&worker->entry) to determine whether destruction or rebinding is pending. This simplifies worker state transitions. WORKER_REBIND is not needed anymore. Remove it. tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: rename manager_mutex to assoc_mutex Now that manager_mutex's role has changed from synchronizing manager role to excluding hotplug against manager, the name is misleading. As it is protecting the CPU-association of the gcwq now, rename it to assoc_mutex. This patch is pure rename and doesn't introduce any functional change. tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use __cpuinit instead of __devinit for cpu callbacks For workqueue hotplug callbacks, it makes less sense to use __devinit which discards the memory after boot if !HOTPLUG. __cpuinit, which discards the memory after boot if !HOTPLUG_CPU fits better. tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix possible stall on try_to_grab_pending() of a delayed work item Currently, when try_to_grab_pending() grabs a delayed work item, it leaves its linked work items alone on the delayed_works. The linked work items are always NO_COLOR and will cause future cwq_activate_first_delayed() increase cwq->nr_active incorrectly, and may cause the whole cwq to stall. For example, state: cwq->max_active = 1, cwq->nr_active = 1 one work in cwq->pool, many in cwq->delayed_works. step1: try_to_grab_pending() removes a work item from delayed_works but leaves its NO_COLOR linked work items on it. step2: Later on, cwq_activate_first_delayed() activates the linked work item increasing ->nr_active. step3: cwq->nr_active = 1, but all activated work items of the cwq are NO_COLOR. When they finish, cwq->nr_active will not be decreased due to NO_COLOR, and no further work items will be activated from cwq->delayed_works. the cwq stalls. Fix it by ensuring the target work item is activated before stealing PENDING in try_to_grab_pending(). This ensures that all the linked work items are activated without incorrectly bumping cwq->nr_active. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@kernel.org workqueue: reimplement work_on_cpu() using system_wq The existing work_on_cpu() implementation is hugely inefficient. It creates a new kthread, execute that single function and then let the kthread die on each invocation. Now that system_wq can handle concurrent executions, there's no advantage of doing this. Reimplement work_on_cpu() using system_wq which makes it simpler and way more efficient. stable: While this isn't a fix in itself, it's needed to fix a workqueue related bug in cpufreq/powernow-k8. AFAICS, this shouldn't break other existing users. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Jiri Kosina <jkosina@suse.cz> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Bjorn Helgaas <bhelgaas@google.com> Cc: Len Brown <lenb@kernel.org> Cc: Rafael J. Wysocki <rjw@sisk.pl> Cc: stable@vger.kernel.org workqueue: introduce cwq_set_max_active() helper for thaw_workqueues() Using a helper instead of open code makes thaw_workqueues() clearer. The helper will also be used by the next patch. tj: Slight update to comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use cwq_set_max_active() helper for workqueue_set_max_active() workqueue_set_max_active() may increase ->max_active without activating delayed works and may make the activation order differ from the queueing order. Both aren't strictly bugs but the resulting behavior could be a bit odd. To make things more consistent, use cwq_set_max_active() helper which immediately makes use of the newly increased max_mactive if there are delayed work items and also keeps the activation order. tj: Slight update to description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: remove spurious WARN_ON_ONCE(in_irq()) from try_to_grab_pending() e0aecdd874 ("workqueue: use irqsafe timer for delayed_work") made try_to_grab_pending() safe to use from irq context but forgot to remove WARN_ON_ONCE(in_irq()). Remove it. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Fengguang Wu <fengguang.wu@intel.com> workqueue: cancel_delayed_work() should return %false if work item is idle 57b30ae77b ("workqueue: reimplement cancel_delayed_work() using try_to_grab_pending()") made cancel_delayed_work() always return %true unless someone else is also trying to cancel the work item, which is broken - if the target work item is idle, the return value should be %false. try_to_grab_pending() indicates that the target work item was idle by zero return value. Use it for return. Note that this brings cancel_delayed_work() in line with __cancel_work_timer() in return value handling. Signed-off-by: Dan Magenheimer <dan.magenheimer@oracle.com> Signed-off-by: Tejun Heo <tj@kernel.org> LKML-Reference: <444a6439-b1a4-4740-9e7e-bc37267cfe73@default> workqueue: exit rescuer_thread() as TASK_RUNNING A rescue thread exiting TASK_INTERRUPTIBLE can lead to a task scheduling off, never to be seen again. In the case where this occurred, an exiting thread hit reiserfs homebrew conditional resched while holding a mutex, bringing the box to its knees. PID: 18105 TASK: ffff8807fd412180 CPU: 5 COMMAND: "kdmflush" #0 [ffff8808157e7670] schedule at ffffffff8143f489 #1 [ffff8808157e77b8] reiserfs_get_block at ffffffffa038ab2d [reiserfs] #2 [ffff8808157e79a8] __block_write_begin at ffffffff8117fb14 #3 [ffff8808157e7a98] reiserfs_write_begin at ffffffffa0388695 [reiserfs] #4 [ffff8808157e7ad8] generic_perform_write at ffffffff810ee9e2 #5 [ffff8808157e7b58] generic_file_buffered_write at ffffffff810eeb41 #6 [ffff8808157e7ba8] __generic_file_aio_write at ffffffff810f1a3a #7 [ffff8808157e7c58] generic_file_aio_write at ffffffff810f1c88 #8 [ffff8808157e7cc8] do_sync_write at ffffffff8114f850 #9 [ffff8808157e7dd8] do_acct_process at ffffffff810a268f [exception RIP: kernel_thread_helper] RIP: ffffffff8144a5c0 RSP: ffff8808157e7f58 RFLAGS: 00000202 RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffffffff8107af60 RDI: ffff8803ee491d18 RBP: 0000000000000000 R8: 0000000000000000 R9: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 Signed-off-by: Mike Galbraith <mgalbraith@suse.de> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org workqueue: mod_delayed_work_on() shouldn't queue timer on 0 delay 8376fe22c7 ("workqueue: implement mod_delayed_work[_on]()") implemented mod_delayed_work[_on]() using the improved try_to_grab_pending(). The function is later used, among others, to replace [__]candel_delayed_work() + queue_delayed_work() combinations. Unfortunately, a delayed_work item w/ zero @delay is handled slightly differently by mod_delayed_work_on() compared to queue_delayed_work_on(). The latter skips timer altogether and directly queues it using queue_work_on() while the former schedules timer which will expire on the closest tick. This means, when @delay is zero, that [__]cancel_delayed_work() + queue_delayed_work_on() makes the target item immediately executable while mod_delayed_work_on() may induce delay of upto a full tick. This somewhat subtle difference breaks some of the converted users. e.g. block queue plugging uses delayed_work for deferred processing and uses mod_delayed_work_on() when the queue needs to be immediately unplugged. The above problem manifested as noticeably higher number of context switches under certain circumstances. The difference in behavior was caused by missing special case handling for 0 delay in mod_delayed_work_on() compared to queue_delayed_work_on(). Joonsoo Kim posted a patch to add it - ("workqueue: optimize mod_delayed_work_on() when @delay == 0")[1]. The patch was queued for 3.8 but it was described as optimization and I missed that it was a correctness issue. As both queue_delayed_work_on() and mod_delayed_work_on() use __queue_delayed_work() for queueing, it seems that the better approach is to move the 0 delay special handling to the function instead of duplicating it in mod_delayed_work_on(). Fix the problem by moving 0 delay special case handling from queue_delayed_work_on() to __queue_delayed_work(). This replaces Joonsoo's patch. [1] http://thread.gmane.org/gmane.linux.kernel/1379011/focus=1379012 Signed-off-by: Tejun Heo <tj@kernel.org> Reported-and-tested-by: Anders Kaseorg <andersk@MIT.EDU> Reported-and-tested-by: Zlatko Calusic <zlatko.calusic@iskon.hr> LKML-Reference: <alpine.DEB.2.00.1211280953350.26602@dr-wily.mit.edu> LKML-Reference: <50A78AA9.5040904@iskon.hr> Cc: Joonsoo Kim <js1304@gmail.com> workqueue: trivial fix for return statement in work_busy() Return type of work_busy() is unsigned int. There is return statement returning boolean value, 'false' in work_busy(). It is not problem, because 'false' may be treated '0'. However, fixing it would make code robust. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: add WARN_ON_ONCE() on CPU number to wq_worker_waking_up() Recently, workqueue code has gone through some changes and we found some bugs related to concurrency management operations happening on the wrong CPU. When a worker is concurrency managed (!WORKER_NOT_RUNNIG), it should be bound to its associated cpu and woken up to that cpu. Add WARN_ON_ONCE() to verify this. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: convert BUG_ON()s in __queue_delayed_work() to WARN_ON_ONCE()s 8852aac25e ("workqueue: mod_delayed_work_on() shouldn't queue timer on 0 delay") unexpectedly uncovered a very nasty abuse of delayed_work in megaraid - it allocated work_struct, casted it to delayed_work and then pass that into queue_delayed_work(). Previously, this was okay because 0 @delay short-circuited to queue_work() before doing anything with delayed_work. 8852aac25e moved 0 @delay test into __queue_delayed_work() after sanity check on delayed_work making megaraid trigger BUG_ON(). Although megaraid is already fixed by c1d390d8e6 ("megaraid: fix BUG_ON() from incorrect use of delayed work"), this patch converts BUG_ON()s in __queue_delayed_work() to WARN_ON_ONCE()s so that such abusers, if there are more, trigger warning but don't crash the machine. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Xiaotian Feng <xtfeng@gmail.com> wq Change-Id: Ia3c507777a995f32bf6b40dc8318203e53134229 Signed-off-by: franciscofranco <franciscofranco.1990@gmail.com>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 20, 2016
This moves ARM over to the asm-generic/unaligned.h header. This has the benefit of better code generated especially for ARMv7 on gcc 4.7+ compilers. As Arnd Bergmann, points out: The asm-generic version uses the "struct" version for native-endian unaligned access and the "byteshift" version for the opposite endianess. The current ARM version however uses the "byteshift" implementation for both. Thanks to Nicolas Pitre for the excellent analysis: Test case: int foo (int *x) { return get_unaligned(x); } long long bar (long long *x) { return get_unaligned(x); } With the current ARM version: foo: ldrb r3, [r0, stratosk#2] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 2B], MEM[(const u8 *)x_1(D) + 2B] ldrb r1, [r0, stratosk#1] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 1B], MEM[(const u8 *)x_1(D) + 1B] ldrb r2, [r0, #0] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D)], MEM[(const u8 *)x_1(D)] mov r3, r3, asl #16 @ tmp154, MEM[(const u8 *)x_1(D) + 2B], ldrb r0, [r0, stratosk#3] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 3B], MEM[(const u8 *)x_1(D) + 3B] orr r3, r3, r1, asl aosp-mirror#8 @, tmp155, tmp154, MEM[(const u8 *)x_1(D) + 1B], orr r3, r3, r2 @ tmp157, tmp155, MEM[(const u8 *)x_1(D)] orr r0, r3, r0, asl #24 @,, tmp157, MEM[(const u8 *)x_1(D) + 3B], bx lr @ bar: stmfd sp!, {r4, r5, r6, r7} @, mov r2, #0 @ tmp184, ldrb r5, [r0, aosp-mirror#6] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 6B], MEM[(const u8 *)x_1(D) + 6B] ldrb r4, [r0, aosp-mirror#5] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 5B], MEM[(const u8 *)x_1(D) + 5B] ldrb ip, [r0, stratosk#2] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 2B], MEM[(const u8 *)x_1(D) + 2B] ldrb r1, [r0, aosp-mirror#4] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 4B], MEM[(const u8 *)x_1(D) + 4B] mov r5, r5, asl #16 @ tmp175, MEM[(const u8 *)x_1(D) + 6B], ldrb r7, [r0, stratosk#1] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 1B], MEM[(const u8 *)x_1(D) + 1B] orr r5, r5, r4, asl aosp-mirror#8 @, tmp176, tmp175, MEM[(const u8 *)x_1(D) + 5B], ldrb r6, [r0, aosp-mirror#7] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 7B], MEM[(const u8 *)x_1(D) + 7B] orr r5, r5, r1 @ tmp178, tmp176, MEM[(const u8 *)x_1(D) + 4B] ldrb r4, [r0, #0] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D)], MEM[(const u8 *)x_1(D)] mov ip, ip, asl #16 @ tmp188, MEM[(const u8 *)x_1(D) + 2B], ldrb r1, [r0, stratosk#3] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 3B], MEM[(const u8 *)x_1(D) + 3B] orr ip, ip, r7, asl aosp-mirror#8 @, tmp189, tmp188, MEM[(const u8 *)x_1(D) + 1B], orr r3, r5, r6, asl #24 @,, tmp178, MEM[(const u8 *)x_1(D) + 7B], orr ip, ip, r4 @ tmp191, tmp189, MEM[(const u8 *)x_1(D)] orr ip, ip, r1, asl #24 @, tmp194, tmp191, MEM[(const u8 *)x_1(D) + 3B], mov r1, r3 @, orr r0, r2, ip @ tmp171, tmp184, tmp194 ldmfd sp!, {r4, r5, r6, r7} bx lr In both cases the code is slightly suboptimal. One may wonder why wasting r2 with the constant 0 in the second case for example. And all the mov's could be folded in subsequent orr's, etc. Now with the asm-generic version: foo: ldr r0, [r0, #0] @ unaligned @,* x bx lr @ bar: mov r3, r0 @ x, x ldr r0, [r0, #0] @ unaligned @,* x ldr r1, [r3, aosp-mirror#4] @ unaligned @, bx lr @ This is way better of course, but only because this was compiled for ARMv7. In this case the compiler knows that the hardware can do unaligned word access. This isn't that obvious for foo(), but if we remove the get_unaligned() from bar as follows: long long bar (long long *x) {return *x; } then the resulting code is: bar: ldmia r0, {r0, r1} @ x,, bx lr @ So this proves that the presumed aligned vs unaligned cases does have influence on the instructions the compiler may use and that the above unaligned code results are not just an accident. Still... this isn't fully conclusive without at least looking at the resulting assembly fron a pre ARMv6 compilation. Let's see with an ARMv5 target: foo: ldrb r3, [r0, #0] @ zero_extendqisi2 @ tmp139,* x ldrb r1, [r0, stratosk#1] @ zero_extendqisi2 @ tmp140, ldrb r2, [r0, stratosk#2] @ zero_extendqisi2 @ tmp143, ldrb r0, [r0, stratosk#3] @ zero_extendqisi2 @ tmp146, orr r3, r3, r1, asl aosp-mirror#8 @, tmp142, tmp139, tmp140, orr r3, r3, r2, asl #16 @, tmp145, tmp142, tmp143, orr r0, r3, r0, asl #24 @,, tmp145, tmp146, bx lr @ bar: stmfd sp!, {r4, r5, r6, r7} @, ldrb r2, [r0, #0] @ zero_extendqisi2 @ tmp139,* x ldrb r7, [r0, stratosk#1] @ zero_extendqisi2 @ tmp140, ldrb r3, [r0, aosp-mirror#4] @ zero_extendqisi2 @ tmp149, ldrb r6, [r0, aosp-mirror#5] @ zero_extendqisi2 @ tmp150, ldrb r5, [r0, stratosk#2] @ zero_extendqisi2 @ tmp143, ldrb r4, [r0, aosp-mirror#6] @ zero_extendqisi2 @ tmp153, ldrb r1, [r0, aosp-mirror#7] @ zero_extendqisi2 @ tmp156, ldrb ip, [r0, stratosk#3] @ zero_extendqisi2 @ tmp146, orr r2, r2, r7, asl aosp-mirror#8 @, tmp142, tmp139, tmp140, orr r3, r3, r6, asl aosp-mirror#8 @, tmp152, tmp149, tmp150, orr r2, r2, r5, asl #16 @, tmp145, tmp142, tmp143, orr r3, r3, r4, asl #16 @, tmp155, tmp152, tmp153, orr r0, r2, ip, asl #24 @,, tmp145, tmp146, orr r1, r3, r1, asl #24 @,, tmp155, tmp156, ldmfd sp!, {r4, r5, r6, r7} bx lr Compared to the initial results, this is really nicely optimized and I couldn't do much better if I were to hand code it myself. Signed-off-by: Rob Herring <rob.herring@calxeda.com> Reviewed-by: Nicolas Pitre <nico@linaro.org> Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> modified for Mako from kernel.org reference Signed-off-by: faux123 <reioux@gmail.com>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 21, 2016
With the introduction of fair queued rwlock, recursive read_lock() may hang the offending process if there is a write_lock() somewhere in between. With recursive read_lock checking enabled, the following error was reported: ============================================= [ INFO: possible recursive locking detected ] 3.16.0-rc1 stratosk#2 Tainted: G E --------------------------------------------- load_policy/708 is trying to acquire lock: (policy_rwlock){.+.+..}, at: [<ffffffff8125b32a>] security_genfs_sid+0x3a/0x170 but task is already holding lock: (policy_rwlock){.+.+..}, at: [<ffffffff8125b48c>] security_fs_use+0x2c/0x110 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(policy_rwlock); lock(policy_rwlock); This patch fixes the occurrence of recursive read_lock() of policy_rwlock by adding a helper function __security_genfs_sid() which requires caller to take the lock before calling it. The security_fs_use() was then modified to call the new helper function. Signed-off-by: Waiman Long <Waiman.Long@hp.com> Acked-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Paul Moore <pmoore@redhat.com> Signed-off-by: Pranav Vashi <neobuddy89@gmail.com>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 21, 2016
msm_sat_enqueue() calls spin_lock() and msm_sat_dequeue() calls spin_lock_irqsave(). This leads to lockdep warnings about the same lock being taken in interrupts on and interrupts off context which can lead to a potential deadlock. ================================= [ INFO: inconsistent lock state ] 3.4.0+ #382 Tainted: G W --------------------------------- inconsistent {HARDIRQ-ON-W} -> {IN-HARDIRQ-W} usage. kworker/u:2/94 [HC1[1]:SC0[0]:HE0:SE1] takes: (&(&sat->lock)->rlock){?.+...}, at: [<c0348b80>] msm_sat_enqueue+0x20/0xb0 {HARDIRQ-ON-W} state was registered at: [<c00bd290>] __lock_acquire+0x664/0x8d8 [<c00bd690>] lock_acquire+0x18c/0x1e8 [<c06989a4>] _raw_spin_lock+0x38/0x48 [<c0348b80>] msm_sat_enqueue+0x20/0xb0 [<c0349338>] msm_slim_rxwq+0x278/0x42c [<c0349580>] msm_slim_rx_msgq_thread+0x94/0x1f8 [<c008e480>] kthread+0x90/0xa0 [<c000f438>] kernel_thread_exit+0x0/0x8 irq event stamp: 24219 hardirqs last enabled at (24218): [<c0699188>] _raw_spin_unlock_irqrestore+0x3c/0x68 hardirqs last disabled at (24219): [<c06993f4>] __irq_svc+0x34/0x78 softirqs last enabled at (24129): [<c00738a8>] irq_exit+0x54/0xa8 softirqs last disabled at (24108): [<c00738a8>] irq_exit+0x54/0xa8 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&(&sat->lock)->rlock); <Interrupt> lock(&(&sat->lock)->rlock); *** DEADLOCK *** 3 locks held by kworker/u:2/94: #0: ((sat->satcl.name)){.+.+.+}, at: [<c0087124>] process_one_work+0x1e8/0x648 stratosk#1: ((&sat->wd)){+.+.+.}, at: [<c0087124>] process_one_work+0x1e8/0x648 stratosk#2: (&dev->tx_lock){+.+...}, at: [<c0349cac>] msm_xfer_msg+0xb4/0x51c stack backtrace: [<c00151b0>] (unwind_backtrace+0x0/0x120) from [<c00b9d48>] (print_usage_bug+0x258/0x2c0) [<c00b9d48>] (print_usage_bug+0x258/0x2c0) from [<c00ba12c>] (mark_lock+0x37c/0x68c) [<c00ba12c>] (mark_lock+0x37c/0x68c) from [<c00bd20c>] (__lock_acquire+0x5e0/0x8d8) [<c00bd20c>] (__lock_acquire+0x5e0/0x8d8) from [<c00bd690>] (lock_acquire+0x18c/0x1e8) [<c00bd690>] (lock_acquire+0x18c/0x1e8) from [<c06989a4>] (_raw_spin_lock+0x38/0x48) [<c06989a4>] (_raw_spin_lock+0x38/0x48) from [<c0348b80>] (msm_sat_enqueue+0x20/0xb0) [<c0348b80>] (msm_sat_enqueue+0x20/0xb0) from [<c0348dc8>] (msm_slim_interrupt+0x1b8/0x4b0) [<c0348dc8>] (msm_slim_interrupt+0x1b8/0x4b0) from [<c00d2ccc>] (handle_irq_event_percpu+0x118/0x3a4) [<c00d2ccc>] (handle_irq_event_percpu+0x118/0x3a4) from [<c00d2f94>] (handle_irq_event+0x3c/0x5c) [<c00d2f94>] (handle_irq_event+0x3c/0x5c) from [<c00d5c4c>] (handle_fasteoi_irq+0xd0/0x11c) [<c00d5c4c>] (handle_fasteoi_irq+0xd0/0x11c) from [<c00d2ae4>] (generic_handle_irq+0x24/0x2c) [<c00d2ae4>] (generic_handle_irq+0x24/0x2c) from [<c000f370>] (handle_IRQ+0x7c/0xc0) [<c000f370>] (handle_IRQ+0x7c/0xc0) from [<c0008774>] (gic_handle_irq+0x6c/0xc4) [<c0008774>] (gic_handle_irq+0x6c/0xc4) from [<c0699404>] (__irq_svc+0x44/0x78) Exception stack(0xee061cd8 to 0xee061d20) 1cc0: 00000001 eebda7c8 1ce0: 00000000 eebda400 20000013 c12b5430 ec96c168 c12b542c c12b5430 00000001 1d00: 20000013 ee061e14 3eb13eb1 ee061d20 c00ba520 c069918c 20000013 ffffffff [<c0699404>] (__irq_svc+0x44/0x78) from [<c069918c>] (_raw_spin_unlock_irqrestore+0x40/0x68) [<c069918c>] (_raw_spin_unlock_irqrestore+0x40/0x68) from [<c0279f2c>] (__debug_object_init+0x30c/0x344) [<c0279f2c>] (__debug_object_init+0x30c/0x344) from [<c007b068>] (init_timer_on_stack_key+0x18/0x30) [<c007b068>] (init_timer_on_stack_key+0x18/0x30) from [<c0695248>] (schedule_timeout+0x8c/0x4c0) [<c0695248>] (schedule_timeout+0x8c/0x4c0) from [<c069768c>] (wait_for_common+0xec/0x164) [<c069768c>] (wait_for_common+0xec/0x164) from [<c0349fc4>] (msm_xfer_msg+0x3cc/0x51c) [<c0349fc4>] (msm_xfer_msg+0x3cc/0x51c) from [<c034a358>] (slim_sat_rxprocess+0x244/0x664) [<c034a358>] (slim_sat_rxprocess+0x244/0x664) from [<c0087290>] (process_one_work+0x354/0x648) [<c0087290>] (process_one_work+0x354/0x648) from [<c0089754>] (worker_thread+0x1a8/0x2a8) [<c0089754>] (worker_thread+0x1a8/0x2a8) from [<c008e480>] (kthread+0x90/0xa0) [<c008e480>] (kthread+0x90/0xa0) from [<c000f438>] (kernel_thread_exit+0x0/0x8) Make this lock irqsafe as well so that this potential bug doesn't occur. Change-Id: Icbef6d1d749ee6ee81b079e19e57f22c38f00c68 Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 21, 2016
workqueue: change BUG_ON() to WARN_ON() This BUG_ON() can be triggered if you call schedule_work() before calling INIT_WORK(). It is a bug definitely, but it's nicer to just print a stack trace and return. Reported-by: Matt Renzelmann <mjr@cs.wisc.edu> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: Catch more locking problems with flush_work() If a workqueue is flushed with flush_work() lockdep checking can be circumvented. For example: static DEFINE_MUTEX(mutex); static void my_work(struct work_struct *w) { mutex_lock(&mutex); mutex_unlock(&mutex); } static DECLARE_WORK(work, my_work); static int __init start_test_module(void) { schedule_work(&work); return 0; } module_init(start_test_module); static void __exit stop_test_module(void) { mutex_lock(&mutex); flush_work(&work); mutex_unlock(&mutex); } module_exit(stop_test_module); would not always print a warning when flush_work() was called. In this trivial example nothing could go wrong since we are guaranteed module_init() and module_exit() don't run concurrently, but if the work item is schedule asynchronously we could have a scenario where the work item is running just at the time flush_work() is called resulting in a classic ABBA locking problem. Add a lockdep hint by acquiring and releasing the work item lockdep_map in flush_work() so that we always catch this potential deadlock scenario. Signed-off-by: Stephen Boyd <sboyd@codeaurora.org> Reviewed-by: Yong Zhang <yong.zhang0@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> lockdep: fix oops in processing workqueue Under memory load, on x86_64, with lockdep enabled, the workqueue's process_one_work() has been seen to oops in __lock_acquire(), barfing on a 0xffffffff00000000 pointer in the lockdep_map's class_cache[]. Because it's permissible to free a work_struct from its callout function, the map used is an onstack copy of the map given in the work_struct: and that copy is made without any locking. Surprisingly, gcc (4.5.1 in Hugh's case) uses "rep movsl" rather than "rep movsq" for that structure copy: which might race with a workqueue user's wait_on_work() doing lock_map_acquire() on the source of the copy, putting a pointer into the class_cache[], but only in time for the top half of that pointer to be copied to the destination map. Boom when process_one_work() subsequently does lock_map_acquire() on its onstack copy of the lockdep_map. Fix this, and a similar instance in call_timer_fn(), with a lockdep_copy_map() function which additionally NULLs the class_cache[]. Note: this oops was actually seen on 3.4-next, where flush_work() newly does the racing lock_map_acquire(); but Tejun points out that 3.4 and earlier are already vulnerable to the same through wait_on_work(). * Patch orginally from Peter. Hugh modified it a bit and wrote the description. Signed-off-by: Peter Zijlstra <peterz@infradead.org> Reported-by: Hugh Dickins <hughd@google.com> LKML-Reference: <alpine.LSU.2.00.1205070951170.1544@eggly.anvils> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: perform cpu down operations from low priority cpu_notifier() Currently, all workqueue cpu hotplug operations run off CPU_PRI_WORKQUEUE which is higher than normal notifiers. This is to ensure that workqueue is up and running while bringing up a CPU before other notifiers try to use workqueue on the CPU. Per-cpu workqueues are supposed to remain working and bound to the CPU for normal CPU_DOWN_PREPARE notifiers. This holds mostly true even with workqueue offlining running with higher priority because workqueue CPU_DOWN_PREPARE only creates a bound trustee thread which runs the per-cpu workqueue without concurrency management without explicitly detaching the existing workers. However, if the trustee needs to create new workers, it creates unbound workers which may wander off to other CPUs while CPU_DOWN_PREPARE notifiers are in progress. Furthermore, if the CPU down is cancelled, the per-CPU workqueue may end up with workers which aren't bound to the CPU. While reliably reproducible with a convoluted artificial test-case involving scheduling and flushing CPU burning work items from CPU down notifiers, this isn't very likely to happen in the wild, and, even when it happens, the effects are likely to be hidden by the following successful CPU down. Fix it by using different priorities for up and down notifiers - high priority for up operations and low priority for down operations. Workqueue cpu hotplug operations will soon go through further cleanup. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: drop CPU_DYING notifier operation Workqueue used CPU_DYING notification to mark GCWQ_DISASSOCIATED. This was necessary because workqueue's CPU_DOWN_PREPARE happened before other DOWN_PREPARE notifiers and workqueue needed to stay associated across the rest of DOWN_PREPARE. After the previous patch, workqueue's DOWN_PREPARE happens after others and can set GCWQ_DISASSOCIATED directly. Drop CPU_DYING and let the trustee set GCWQ_DISASSOCIATED after disabling concurrency management. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: ROGUE workers are UNBOUND workers Currently, WORKER_UNBOUND is used to mark workers for the unbound global_cwq and WORKER_ROGUE is used to mark workers for disassociated per-cpu global_cwqs. Both are used to make the marked worker skip concurrency management and the only place they make any difference is in worker_enter_idle() where WORKER_ROGUE is used to skip scheduling idle timer, which can easily be replaced with trustee state testing. This patch replaces WORKER_ROGUE with WORKER_UNBOUND and drops WORKER_ROGUE. This is to prepare for removing trustee and handling disassociated global_cwqs as unbound. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: use mutex for global_cwq manager exclusion POOL_MANAGING_WORKERS is used to ensure that at most one worker takes the manager role at any given time on a given global_cwq. Trustee later hitched on it to assume manager adding blocking wait for the bit. As trustee already needed a custom wait mechanism, waiting for MANAGING_WORKERS was rolled into the same mechanism. Trustee is scheduled to be removed. This patch separates out MANAGING_WORKERS wait into per-pool mutex. Workers use mutex_trylock() to test for manager role and trustee uses mutex_lock() to claim manager roles. gcwq_claim/release_management() helpers are added to grab and release manager roles of all pools on a global_cwq. gcwq_claim_management() always grabs pool manager mutexes in ascending pool index order and uses pool index as lockdep subclass. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: drop @bind from create_worker() Currently, create_worker()'s callers are responsible for deciding whether the newly created worker should be bound to the associated CPU and create_worker() sets WORKER_UNBOUND only for the workers for the unbound global_cwq. Creation during normal operation is always via maybe_create_worker() and @bind is true. For workers created during hotplug, @bind is false. Normal operation path is planned to be used even while the CPU is going through hotplug operations or offline and this static decision won't work. Drop @bind from create_worker() and decide whether to bind by looking at GCWQ_DISASSOCIATED. create_worker() will also set WORKER_UNBOUND autmatically if disassociated. To avoid flipping GCWQ_DISASSOCIATED while create_worker() is in progress, the flag is now allowed to be changed only while holding all manager_mutexes on the global_cwq. This requires that GCWQ_DISASSOCIATED is not cleared behind trustee's back. CPU_ONLINE no longer clears DISASSOCIATED before flushing trustee, which clears DISASSOCIATED before rebinding remaining workers if asked to release. For cases where trustee isn't around, CPU_ONLINE clears DISASSOCIATED after flushing trustee. Also, now, first_idle has UNBOUND set on creation which is explicitly cleared by CPU_ONLINE while binding it. These convolutions will soon be removed by further simplification of CPU hotplug path. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: reimplement CPU online rebinding to handle idle workers Currently, if there are left workers when a CPU is being brough back online, the trustee kills all idle workers and scheduled rebind_work so that they re-bind to the CPU after the currently executing work is finished. This works for busy workers because concurrency management doesn't try to wake up them from scheduler callbacks, which require the target task to be on the local run queue. The busy worker bumps concurrency counter appropriately as it clears WORKER_UNBOUND from the rebind work item and it's bound to the CPU before returning to the idle state. To reduce CPU on/offlining overhead (as many embedded systems use it for powersaving) and simplify the code path, workqueue is planned to be modified to retain idle workers across CPU on/offlining. This patch reimplements CPU online rebinding such that it can also handle idle workers. As noted earlier, due to the local wakeup requirement, rebinding idle workers is tricky. All idle workers must be re-bound before scheduler callbacks are enabled. This is achieved by interlocking idle re-binding. Idle workers are requested to re-bind and then hold until all idle re-binding is complete so that no bound worker starts executing work item. Only after all idle workers are re-bound and parked, CPU_ONLINE proceeds to release them and queue rebind work item to busy workers thus guaranteeing scheduler callbacks aren't invoked until all idle workers are ready. worker_rebind_fn() is renamed to busy_worker_rebind_fn() and idle_worker_rebind() for idle workers is added. Rebinding logic is moved to rebind_workers() and now called from CPU_ONLINE after flushing trustee. While at it, add CPU sanity check in worker_thread(). Note that now a worker may become idle or the manager between trustee release and rebinding during CPU_ONLINE. As the previous patch updated create_worker() so that it can be used by regular manager while unbound and this patch implements idle re-binding, this is safe. This prepares for removal of trustee and keeping idle workers across CPU hotplugs. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: don't butcher idle workers on an offline CPU Currently, during CPU offlining, after all pending work items are drained, the trustee butchers all workers. Also, on CPU onlining failure, workqueue_cpu_callback() ensures that the first idle worker is destroyed. Combined, these guarantee that an offline CPU doesn't have any worker for it once all the lingering work items are finished. This guarantee isn't really necessary and makes CPU on/offlining more expensive than needs to be, especially for platforms which use CPU hotplug for powersaving. This patch lets offline CPUs removes idle worker butchering from the trustee and let a CPU which failed onlining keep the created first worker. The first worker is created if the CPU doesn't have any during CPU_DOWN_PREPARE and started right away. If onlining succeeds, the rebind_workers() call in CPU_ONLINE will rebind it like any other workers. If onlining fails, the worker is left alone till the next try. This makes CPU hotplugs cheaper by allowing global_cwqs to keep workers across them and simplifies code. Note that trustee doesn't re-arm idle timer when it's done and thus the disassociated global_cwq will keep all workers until it comes back online. This will be improved by further patches. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: remove CPU offline trustee With the previous changes, a disassociated global_cwq now can run as an unbound one on its own - it can create workers as necessary to drain remaining works after the CPU has been brought down and manage the number of workers using the usual idle timer mechanism making trustee completely redundant except for the actual unbinding operation. This patch removes the trustee and let a disassociated global_cwq manage itself. Unbinding is moved to a work item (for CPU affinity) which is scheduled and flushed from CPU_DONW_PREPARE. This patch moves nr_running clearing outside gcwq and manager locks to simplify the code. As nr_running is unused at the point, this is safe. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: simplify CPU hotplug code With trustee gone, CPU hotplug code can be simplified. * gcwq_claim/release_management() now grab and release gcwq lock too respectively and gained _and_lock and _and_unlock postfixes. * All CPU hotplug logic was implemented in workqueue_cpu_callback() which was called by workqueue_cpu_up/down_callback() for the correct priority. This was because up and down paths shared a lot of logic, which is no longer true. Remove workqueue_cpu_callback() and move all hotplug logic into the two actual callbacks. This patch doesn't make any functional changes. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: fix spurious CPU locality WARN from process_one_work() 25511a4776 "workqueue: reimplement CPU online rebinding to handle idle workers" added CPU locality sanity check in process_one_work(). It triggers if a worker is executing on a different CPU without UNBOUND or REBIND set. This works for all normal workers but rescuers can trigger this spuriously when they're serving the unbound or a disassociated global_cwq - rescuers don't have either flag set and thus its gcwq->cpu can be a different value including %WORK_CPU_UNBOUND. Fix it by additionally testing %GCWQ_DISASSOCIATED. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> LKML-Refence: <20120721213656.GA7783@linux.vnet.ibm.com> workqueue: reorder queueing functions so that _on() variants are on top Currently, queue/schedule[_delayed]_work_on() are located below the counterpart without the _on postifx even though the latter is usually implemented using the former. Swap them. This is cleanup and doesn't cause any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: make queueing functions return bool All queueing functions return 1 on success, 0 if the work item was already pending. Update them to return bool instead. This signifies better that they don't return 0 / -errno. This is cleanup and doesn't cause any functional difference. While at it, fix comment opening for schedule_work_on(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: add missing smp_wmb() in process_one_work() WORK_STRUCT_PENDING is used to claim ownership of a work item and process_one_work() releases it before starting execution. When someone else grabs PENDING, all pre-release updates to the work item should be visible and all updates made by the new owner should happen afterwards. Grabbing PENDING uses test_and_set_bit() and thus has a full barrier; however, clearing doesn't have a matching wmb. Given the preceding spin_unlock and use of clear_bit, I don't believe this can be a problem on an actual machine and there hasn't been any related report but it still is theretically possible for clear_pending to permeate upwards and happen before work->entry update. Add an explicit smp_wmb() before work_clear_pending(). Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: stable@vger.kernel.org workqueue: disable irq while manipulating PENDING Queueing operations use WORK_STRUCT_PENDING_BIT to synchronize access to the target work item. They first try to claim the bit and proceed with queueing only after that succeeds and there's a window between PENDING being set and the actual queueing where the task can be interrupted or preempted. There's also a similar window in process_one_work() when clearing PENDING. A work item is dequeued, gcwq->lock is released and then PENDING is cleared and the worker might get interrupted or preempted between releasing gcwq->lock and clearing PENDING. cancel[_delayed]_work_sync() tries to claim or steal PENDING. The function assumes that a work item with PENDING is either queued or in the process of being [de]queued. In the latter case, it busy-loops until either the work item loses PENDING or is queued. If canceling coincides with the above described interrupts or preemptions, the canceling task will busy-loop while the queueing or executing task is preempted. This patch keeps irq disabled across claiming PENDING and actual queueing and moves PENDING clearing in process_one_work() inside gcwq->lock so that busy looping from PENDING && !queued doesn't wait for interrupted/preempted tasks. Note that, in process_one_work(), setting last CPU and clearing PENDING got merged into single operation. This removes possible long busy-loops and will allow using try_to_grab_pending() from bh and irq contexts. v2: __queue_work() was testing preempt_count() to ensure that the caller has disabled preemption. This triggers spuriously if !CONFIG_PREEMPT_COUNT. Use preemptible() instead. Reported by Fengguang Wu. v3: Disable irq instead of preemption. IRQ will be disabled while grabbing gcwq->lock later anyway and this allows using try_to_grab_pending() from bh and irq contexts. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Fengguang Wu <fengguang.wu@intel.com> workqueue: set delayed_work->timer function on initialization delayed_work->timer.function is currently initialized during queue_delayed_work_on(). Export delayed_work_timer_fn() and set delayed_work timer function during delayed_work initialization together with other fields. This ensures the timer function is always valid on an initialized delayed_work. This is to help mod_delayed_work() implementation. To detect delayed_work users which diddle with the internal timer, trigger WARN if timer function doesn't match on queue. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: unify local CPU queueing handling Queueing functions have been using different methods to determine the local CPU. * queue_work() superflously uses get/put_cpu() to acquire and hold the local CPU across queue_work_on(). * delayed_work_timer_fn() uses smp_processor_id(). * queue_delayed_work() calls queue_delayed_work_on() with -1 @cpu which is interpreted as the local CPU. * flush_delayed_work[_sync]() were using raw_smp_processor_id(). * __queue_work() interprets %WORK_CPU_UNBOUND as local CPU if the target workqueue is bound one but nobody uses this. This patch converts all functions to uniformly use %WORK_CPU_UNBOUND to indicate local CPU and use the local binding feature of __queue_work(). unlikely() is dropped from %WORK_CPU_UNBOUND handling in __queue_work(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix zero @delay handling of queue_delayed_work_on() If @delay is zero and the dealyed_work is idle, queue_delayed_work() queues it for immediate execution; however, queue_delayed_work_on() lacks this logic and always goes through timer regardless of @delay. This patch moves 0 @delay handling logic from queue_delayed_work() to queue_delayed_work_on() so that both functions behave the same. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: move try_to_grab_pending() upwards try_to_grab_pending() will be used by to-be-implemented mod_delayed_work[_on](). Move try_to_grab_pending() and related functions above queueing functions. This patch only moves functions around. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: introduce WORK_OFFQ_FLAG_* Low WORK_STRUCT_FLAG_BITS bits of work_struct->data contain WORK_STRUCT_FLAG_* and flush color. If the work item is queued, the rest point to the cpu_workqueue with WORK_STRUCT_CWQ set; otherwise, WORK_STRUCT_CWQ is clear and the bits contain the last CPU number - either a real CPU number or one of WORK_CPU_*. Scheduled addition of mod_delayed_work[_on]() requires an additional flag, which is used only while a work item is off queue. There are more than enough bits to represent off-queue CPU number on both 32 and 64bits. This patch introduces WORK_OFFQ_FLAG_* which occupy the lower part of the @work->data high bits while off queue. This patch doesn't define any actual OFFQ flag yet. Off-queue CPU number is now shifted by WORK_OFFQ_CPU_SHIFT, which adds the number of bits used by OFFQ flags to WORK_STRUCT_FLAG_SHIFT, to make room for OFFQ flags. To avoid shift width warning with large WORK_OFFQ_FLAG_BITS, ulong cast is added to WORK_STRUCT_NO_CPU and, just in case, BUILD_BUG_ON() to check that there are enough bits to accomodate off-queue CPU number is added. This patch doesn't make any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: factor out __queue_delayed_work() from queue_delayed_work_on() This is to prepare for mod_delayed_work[_on]() and doesn't cause any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reorganize try_to_grab_pending() and __cancel_timer_work() * Use bool @is_dwork instead of @timer and let try_to_grab_pending() use to_delayed_work() to determine the delayed_work address. * Move timer handling from __cancel_work_timer() to try_to_grab_pending(). * Make try_to_grab_pending() use -EAGAIN instead of -1 for busy-looping and drop the ret local variable. * Add proper function comment to try_to_grab_pending(). This makes the code a bit easier to understand and will ease further changes. This patch doesn't make any functional change. v2: Use @is_dwork instead of @timer. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: mark a work item being canceled as such There can be two reasons try_to_grab_pending() can fail with -EAGAIN. One is when someone else is queueing or deqeueing the work item. With the previous patches, it is guaranteed that PENDING and queued state will soon agree making it safe to busy-retry in this case. The other is if multiple __cancel_work_timer() invocations are racing one another. __cancel_work_timer() grabs PENDING and then waits for running instances of the target work item on all CPUs while holding PENDING and !queued. try_to_grab_pending() invoked from another task will keep returning -EAGAIN while the current owner is waiting. Not distinguishing the two cases is okay because __cancel_work_timer() is the only user of try_to_grab_pending() and it invokes wait_on_work() whenever grabbing fails. For the first case, busy looping should be fine but wait_on_work() doesn't cause any critical problem. For the latter case, the new contender usually waits for the same condition as the current owner, so no unnecessarily extended busy-looping happens. Combined, these make __cancel_work_timer() technically correct even without irq protection while grabbing PENDING or distinguishing the two different cases. While the current code is technically correct, not distinguishing the two cases makes it difficult to use try_to_grab_pending() for other purposes than canceling because it's impossible to tell whether it's safe to busy-retry grabbing. This patch adds a mechanism to mark a work item being canceled. try_to_grab_pending() now disables irq on success and returns -EAGAIN to indicate that grabbing failed but PENDING and queued states are gonna agree soon and it's safe to busy-loop. It returns -ENOENT if the work item is being canceled and it may stay PENDING && !queued for arbitrary amount of time. __cancel_work_timer() is modified to mark the work canceling with WORK_OFFQ_CANCELING after grabbing PENDING, thus making try_to_grab_pending() fail with -ENOENT instead of -EAGAIN. Also, it invokes wait_on_work() iff grabbing failed with -ENOENT. This isn't necessary for correctness but makes it consistent with other future users of try_to_grab_pending(). v2: try_to_grab_pending() was testing preempt_count() to ensure that the caller has disabled preemption. This triggers spuriously if !CONFIG_PREEMPT_COUNT. Use preemptible() instead. Reported by Fengguang Wu. v3: Updated so that try_to_grab_pending() disables irq on success rather than requiring preemption disabled by the caller. This makes busy-looping easier and will allow try_to_grap_pending() to be used from bh/irq contexts. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Fengguang Wu <fengguang.wu@intel.com> workqueue: implement mod_delayed_work[_on]() Workqueue was lacking a mechanism to modify the timeout of an already pending delayed_work. delayed_work users have been working around this using several methods - using an explicit timer + work item, messing directly with delayed_work->timer, and canceling before re-queueing, all of which are error-prone and/or ugly. This patch implements mod_delayed_work[_on]() which behaves similarly to mod_timer() - if the delayed_work is idle, it's queued with the given delay; otherwise, its timeout is modified to the new value. Zero @delay guarantees immediate execution. v2: Updated to reflect try_to_grab_pending() changes. Now safe to be called from bh context. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> workqueue: fix CPU binding of flush_delayed_work[_sync]() delayed_work encodes the workqueue to use and the last CPU in delayed_work->work.data while it's on timer. The target CPU is implicitly recorded as the CPU the timer is queued on and delayed_work_timer_fn() queues delayed_work->work to the CPU it is running on. Unfortunately, this leaves flush_delayed_work[_sync]() no way to find out which CPU the delayed_work was queued for when they try to re-queue after killing the timer. Currently, it chooses the local CPU flush is running on. This can unexpectedly move a delayed_work queued on a specific CPU to another CPU and lead to subtle errors. There isn't much point in trying to save several bytes in struct delayed_work, which is already close to a hundred bytes on 64bit with all debug options turned off. This patch adds delayed_work->cpu to remember the CPU it's queued for. Note that if the timer is migrated during CPU down, the work item could be queued to the downed global_cwq after this change. As a detached global_cwq behaves like an unbound one, this doesn't change much for the delayed_work. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> workqueue: add missing wmb() in clear_work_data() Any operation which clears PENDING should be preceded by a wmb to guarantee that the next PENDING owner sees all the changes made before PENDING release. There are only two places where PENDING is cleared - set_work_cpu_and_clear_pending() and clear_work_data(). The caller of the former already does smp_wmb() but the latter doesn't have any. Move the wmb above set_work_cpu_and_clear_pending() into it and add one to clear_work_data(). There hasn't been any report related to this issue, and, given how clear_work_data() is used, it is extremely unlikely to have caused any actual problems on any architecture. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> workqueue: use enum value to set array size of pools in gcwq Commit 3270476a6c0ce322354df8679652f060d66526dc ('workqueue: reimplement WQ_HIGHPRI using a separate worker_pool') introduce separate worker_pool for HIGHPRI. Although there is NR_WORKER_POOLS enum value which represent size of pools, definition of worker_pool in gcwq doesn't use it. Using it makes code robust and prevent future mistakes. So change code to use this enum value. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: correct req_cpu in trace_workqueue_queue_work() When we do tracing workqueue_queue_work(), it records requested cpu. But, if !(@wq->flag & WQ_UNBOUND) and @cpu is WORK_CPU_UNBOUND, requested cpu is changed as local cpu. In case of @wq->flag & WQ_UNBOUND, above change is not occured, therefore it is reasonable to correct it. Use temporary local variable for storing requested cpu. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: change value of lcpu in __queue_delayed_work_on() We assign cpu id into work struct's data field in __queue_delayed_work_on(). In current implementation, when work is come in first time, current running cpu id is assigned. If we do __queue_delayed_work_on() with CPU A on CPU B, __queue_work() invoked in delayed_work_timer_fn() go into the following sub-optimal path in case of WQ_NON_REENTRANT. gcwq = get_gcwq(cpu); if (wq->flags & WQ_NON_REENTRANT && (last_gcwq = get_work_gcwq(work)) && last_gcwq != gcwq) { Change lcpu to @cpu and rechange lcpu to local cpu if lcpu is WORK_CPU_UNBOUND. It is sufficient to prevent to go into sub-optimal path. tj: Slightly rephrased the comment. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: introduce system_highpri_wq Commit 3270476a6c0ce322354df8679652f060d66526dc ('workqueue: reimplement WQ_HIGHPRI using a separate worker_pool') introduce separate worker pool for HIGHPRI. When we handle busyworkers for gcwq, it can be normal worker or highpri worker. But, we don't consider this difference in rebind_workers(), we use just system_wq for highpri worker. It makes mismatch between cwq->pool and worker->pool. It doesn't make error in current implementation, but possible in the future. Now, we introduce system_highpri_wq to use proper cwq for highpri workers in rebind_workers(). Following patch fix this issue properly. tj: Even apart from rebinding, having system_highpri_wq generally makes sense. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use system_highpri_wq for highpri workers in rebind_workers() In rebind_workers(), we do inserting a work to rebind to cpu for busy workers. Currently, in this case, we use only system_wq. This makes a possible error situation as there is mismatch between cwq->pool and worker->pool. To prevent this, we should use system_highpri_wq for highpri worker to match theses. This implements it. tj: Rephrased comment a bit. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use system_highpri_wq for unbind_work To speed cpu down processing up, use system_highpri_wq. As scheduling priority of workers on it is higher than system_wq and it is not contended by other normal works on this cpu, work on it is processed faster than system_wq. tj: CPU up/downs care quite a bit about latency these days. This shouldn't hurt anything and makes sense. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix checkpatch issues Fixed some checkpatch warnings. tj: adapted to wq/for-3.7 and massaged pr_xxx() format strings a bit. Signed-off-by: Valentin Ilie <valentin.ilie@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> LKML-Reference: <1345326762-21747-1-git-send-email-valentin.ilie@gmail.com> workqueue: make all workqueues non-reentrant By default, each per-cpu part of a bound workqueue operates separately and a work item may be executing concurrently on different CPUs. The behavior avoids some cross-cpu traffic but leads to subtle weirdities and not-so-subtle contortions in the API. * There's no sane usefulness in allowing a single work item to be executed concurrently on multiple CPUs. People just get the behavior unintentionally and get surprised after learning about it. Most either explicitly synchronize or use non-reentrant/ordered workqueue but this is error-prone. * flush_work() can't wait for multiple instances of the same work item on different CPUs. If a work item is executing on cpu0 and then queued on cpu1, flush_work() can only wait for the one on cpu1. Unfortunately, work items can easily cross CPU boundaries unintentionally when the queueing thread gets migrated. This means that if multiple queuers compete, flush_work() can't even guarantee that the instance queued right before it is finished before returning. * flush_work_sync() was added to work around some of the deficiencies of flush_work(). In addition to the usual flushing, it ensures that all currently executing instances are finished before returning. This operation is expensive as it has to walk all CPUs and at the same time fails to address competing queuer case. Incorrectly using flush_work() when flush_work_sync() is necessary is an easy error to make and can lead to bugs which are difficult to reproduce. * Similar problems exist for flush_delayed_work[_sync](). Other than the cross-cpu access concern, there's no benefit in allowing parallel execution and it's plain silly to have this level of contortion for workqueue which is widely used from core code to extremely obscure drivers. This patch makes all workqueues non-reentrant. If a work item is executing on a different CPU when queueing is requested, it is always queued to that CPU. This guarantees that any given work item can be executing on one CPU at maximum and if a work item is queued and executing, both are on the same CPU. The only behavior change which may affect workqueue users negatively is that non-reentrancy overrides the affinity specified by queue_work_on(). On a reentrant workqueue, the affinity specified by queue_work_on() is always followed. Now, if the work item is executing on one of the CPUs, the work item will be queued there regardless of the requested affinity. I've reviewed all workqueue users which request explicit affinity, and, fortunately, none seems to be crazy enough to exploit parallel execution of the same work item. This adds an additional busy_hash lookup if the work item was previously queued on a different CPU. This shouldn't be noticeable under any sane workload. Work item queueing isn't a very high-frequency operation and they don't jump across CPUs all the time. In a micro benchmark to exaggerate this difference - measuring the time it takes for two work items to repeatedly jump between two CPUs a number (10M) of times with busy_hash table densely populated, the difference was around 3%. While the overhead is measureable, it is only visible in pathological cases and the difference isn't huge. This change brings much needed sanity to workqueue and makes its behavior consistent with timer. I think this is the right tradeoff to make. This enables significant simplification of workqueue API. Simplification patches will follow. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: gut flush[_delayed]_work_sync() Now that all workqueues are non-reentrant, flush[_delayed]_work_sync() are equivalent to flush[_delayed]_work(). Drop the separate implementation and make them thin wrappers around flush[_delayed]_work(). * start_flush_work() no longer takes @wait_executing as the only left user - flush_work() - always sets it to %true. * __cancel_work_timer() uses flush_work() instead of wait_on_work(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: gut system_nrt[_freezable]_wq() Now that all workqueues are non-reentrant, system[_freezable]_wq() are equivalent to system_nrt[_freezable]_wq(). Replace the latter with wrappers around system[_freezable]_wq(). The wrapping goes through inline functions so that __deprecated can be added easily. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: cosmetic whitespace updates for macro definitions Consistently use the last tab position for '\' line continuation in complex macro definitions. This is to help the following patches. This patch is cosmetic. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use hotcpu_notifier() for workqueue_cpu_down_callback() workqueue_cpu_down_callback() is used only if HOTPLUG_CPU=y, so hotcpu_notifier() fits better than cpu_notifier(). When HOTPLUG_CPU=y, hotcpu_notifier() and cpu_notifier() are the same. When HOTPLUG_CPU=n, if we use cpu_notifier(), workqueue_cpu_down_callback() will be called during boot to do nothing, and the memory of workqueue_cpu_down_callback() and gcwq_unbind_fn() will be discarded after boot. If we use hotcpu_notifier(), we can avoid the no-op call of workqueue_cpu_down_callback() and the memory of workqueue_cpu_down_callback() and gcwq_unbind_fn() will be discard at build time: $ ls -l kernel/workqueue.o.cpu_notifier kernel/workqueue.o.hotcpu_notifier -rw-rw-r-- 1 laijs laijs 484080 Sep 15 11:31 kernel/workqueue.o.cpu_notifier -rw-rw-r-- 1 laijs laijs 478240 Sep 15 11:31 kernel/workqueue.o.hotcpu_notifier $ size kernel/workqueue.o.cpu_notifier kernel/workqueue.o.hotcpu_notifier text data bss dec hex filename 18513 2387 1221 22121 5669 kernel/workqueue.o.cpu_notifier 18082 2355 1221 21658 549a kernel/workqueue.o.hotcpu_notifier tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reimplement cancel_delayed_work() using try_to_grab_pending() cancel_delayed_work() can't be called from IRQ handlers due to its use of del_timer_sync() and can't cancel work items which are already transferred from timer to worklist. Also, unlike other flush and cancel functions, a canceled delayed_work would still point to the last associated cpu_workqueue. If the workqueue is destroyed afterwards and the work item is re-used on a different workqueue, the queueing code can oops trying to dereference already freed cpu_workqueue. This patch reimplements cancel_delayed_work() using try_to_grab_pending() and set_work_cpu_and_clear_pending(). This allows the function to be called from IRQ handlers and makes its behavior consistent with other flush / cancel functions. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> workqueue: UNBOUND -> REBIND morphing in rebind_workers() should be atomic The compiler may compile the following code into TWO write/modify instructions. worker->flags &= ~WORKER_UNBOUND; worker->flags |= WORKER_REBIND; so the other CPU may temporarily see worker->flags which doesn't have either WORKER_UNBOUND or WORKER_REBIND set and perform local wakeup prematurely. Fix it by using single explicit assignment via ACCESS_ONCE(). Because idle workers have another WORKER_NOT_RUNNING flag, this bug doesn't exist for them; however, update it to use the same pattern for consistency. tj: Applied the change to idle workers too and updated comments and patch description a bit. Change-Id: I9b95f51d146c40c31ba028668d6f412bd74c6026 Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org workqueue: move WORKER_REBIND clearing in rebind_workers() to the end of the function This doesn't make any functional difference and is purely to help the next patch to be simpler. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> workqueue: fix possible deadlock in idle worker rebinding Currently, rebind_workers() and idle_worker_rebind() are two-way interlocked. rebind_workers() waits for idle workers to finish rebinding and rebound idle workers wait for rebind_workers() to finish rebinding busy workers before proceeding. Unfortunately, this isn't enough. The second wait from idle workers is implemented as follows. wait_event(gcwq->rebind_hold, !(worker->flags & WORKER_REBIND)); rebind_workers() clears WORKER_REBIND, wakes up the idle workers and then returns. If CPU hotplug cycle happens again before one of the idle workers finishes the above wait_event(), rebind_workers() will repeat the first part of the handshake - set WORKER_REBIND again and wait for the idle worker to finish rebinding - and this leads to deadlock because the idle worker would be waiting for WORKER_REBIND to clear. This is fixed by adding another interlocking step at the end - rebind_workers() now waits for all the idle workers to finish the above WORKER_REBIND wait before returning. This ensures that all rebinding steps are complete on all idle workers before the next hotplug cycle can happen. This problem was diagnosed by Lai Jiangshan who also posted a patch to fix the issue, upon which this patch is based. This is the minimal fix and further patches are scheduled for the next merge window to simplify the CPU hotplug path. Signed-off-by: Tejun Heo <tj@kernel.org> Original-patch-by: Lai Jiangshan <laijs@cn.fujitsu.com> LKML-Reference: <1346516916-1991-3-git-send-email-laijs@cn.fujitsu.com> workqueue: restore POOL_MANAGING_WORKERS This patch restores POOL_MANAGING_WORKERS which was replaced by pool->manager_mutex by 6037315269 "workqueue: use mutex for global_cwq manager exclusion". There's a subtle idle worker depletion bug across CPU hotplug events and we need to distinguish an actual manager and CPU hotplug preventing management. POOL_MANAGING_WORKERS will be used for the former and manager_mutex the later. This patch just lays POOL_MANAGING_WORKERS on top of the existing manager_mutex and doesn't introduce any synchronization changes. The next patch will update it. Note that this patch fixes a non-critical anomaly where too_many_workers() may return %true spuriously while CPU hotplug is in progress. While the issue could schedule idle timer spuriously, it didn't trigger any actual misbehavior. tj: Rewrote patch description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix possible idle worker depletion across CPU hotplug To simplify both normal and CPU hotplug paths, worker management is prevented while CPU hoplug is in progress. This is achieved by CPU hotplug holding the same exclusion mechanism used by workers to ensure there's only one manager per pool. If someone else seems to be performing the manager role, workers proceed to execute work items. CPU hotplug using the same mechanism can lead to idle worker depletion because all workers could proceed to execute work items while CPU hotplug is in progress and CPU hotplug itself wouldn't actually perform the worker management duty - it doesn't guarantee that there's an idle worker left when it releases management. This idle worker depletion, under extreme circumstances, can break forward-progress guarantee and thus lead to deadlock. This patch fixes the bug by using separate mechanisms for manager exclusion among workers and hotplug exclusion. For manager exclusion, POOL_MANAGING_WORKERS which was restored by the previous patch is used. pool->manager_mutex is now only used for exclusion between the elected manager and CPU hotplug. The elected manager won't proceed without holding pool->manager_mutex. This ensures that the worker which won the manager position can't skip managing while CPU hotplug is in progress. It will block on manager_mutex and perform management after CPU hotplug is complete. Note that hotplug may happen while waiting for manager_mutex. A manager isn't either on idle or busy list and thus the hoplug code can't unbind/rebind it. Make the manager handle its own un/rebinding. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: always clear WORKER_REBIND in busy_worker_rebind_fn() busy_worker_rebind_fn() didn't clear WORKER_REBIND if rebinding failed (CPU is down again). This used to be okay because the flag wasn't used for anything else. However, after 25511a477 "workqueue: reimplement CPU online rebinding to handle idle workers", WORKER_REBIND is also used to command idle workers to rebind. If not cleared, the worker may confuse the next CPU_UP cycle by having REBIND spuriously set or oops / get stuck by prematurely calling idle_worker_rebind(). WARNING: at /work/os/wq/kernel/workqueue.c:1323 worker_thread+0x4cd/0x5 00() Hardware name: Bochs Modules linked in: test_wq(O-) Pid: 33, comm: kworker/1:1 Tainted: G O 3.6.0-rc1-work+ #3 Call Trace: [<ffffffff8109039f>] warn_slowpath_common+0x7f/0xc0 [<ffffffff810903fa>] warn_slowpath_null+0x1a/0x20 [<ffffffff810b3f1d>] worker_thread+0x4cd/0x500 [<ffffffff810bc16e>] kthread+0xbe/0xd0 [<ffffffff81bd2664>] kernel_thread_helper+0x4/0x10 ---[ end trace e977cf20f4661968 ]--- BUG: unable to handle kernel NULL pointer dereference at (null) IP: [<ffffffff810b3db0>] worker_thread+0x360/0x500 PGD 0 Oops: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC Modules linked in: test_wq(O-) CPU 0 Pid: 33, comm: kworker/1:1 Tainted: G W O 3.6.0-rc1-work+ #3 Bochs Bochs RIP: 0010:[<ffffffff810b3db0>] [<ffffffff810b3db0>] worker_thread+0x360/0x500 RSP: 0018:ffff88001e1c9de0 EFLAGS: 00010086 RAX: 0000000000000000 RBX: ffff88001e633e00 RCX: 0000000000004140 RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000009 RBP: ffff88001e1c9ea0 R08: 0000000000000000 R09: 0000000000000001 R10: 0000000000000002 R11: 0000000000000000 R12: ffff88001fc8d580 R13: ffff88001fc8d590 R14: ffff88001e633e20 R15: ffff88001e1c6900 FS: 0000000000000000(0000) GS:ffff88001fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 0000000000000000 CR3: 00000000130e8000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process kworker/1:1 (pid: 33, threadinfo ffff88001e1c8000, task ffff88001e1c6900) Stack: ffff880000000000 ffff88001e1c9e40 0000000000000001 ffff88001e1c8010 ffff88001e519c78 ffff88001e1c9e58 ffff88001e1c6900 ffff88001e1c6900 ffff88001e1c6900 ffff88001e1c6900 ffff88001fc8d340 ffff88001fc8d340 Call Trace: [<ffffffff810bc16e>] kthread+0xbe/0xd0 [<ffffffff81bd2664>] kernel_thread_helper+0x4/0x10 Code: b1 00 f6 43 48 02 0f 85 91 01 00 00 48 8b 43 38 48 89 df 48 8b 00 48 89 45 90 e8 ac f0 ff ff 3c 01 0f 85 60 01 00 00 48 8b 53 50 <8b> 02 83 e8 01 85 c0 89 02 0f 84 3b 01 00 00 48 8b 43 38 48 8b RIP [<ffffffff810b3db0>] worker_thread+0x360/0x500 RSP <ffff88001e1c9de0> CR2: 0000000000000000 There was no reason to keep WORKER_REBIND on failure in the first place - WORKER_UNBOUND is guaranteed to be set in such cases preventing incorrectly activating concurrency management. Always clear WORKER_REBIND. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reimplement idle worker rebinding Currently rebind_workers() uses rebinds idle workers synchronously before proceeding to requesting busy workers to rebind. This is necessary because all workers on @worker_pool->idle_list must be bound before concurrency management local wake-ups from the busy workers take place. Unfortunately, the synchronous idle rebinding is quite complicated. This patch reimplements idle rebinding to simplify the code path. Rather than trying to make all idle workers bound before rebinding busy workers, we simply remove all to-be-bound idle workers from the idle list and let them add themselves back after completing rebinding (successful or not). As only workers which finished rebinding can on on the idle worker list, the idle worker list is guaranteed to have only bound workers unless CPU went down again and local wake-ups are safe. After the change, @worker_pool->nr_idle may deviate than the actual number of idle workers on @worker_pool->idle_list. More specifically, nr_idle may be non-zero while ->idle_list is empty. All users of ->nr_idle and ->idle_list are audited. The only affected one is too_many_workers() which is updated to check %false if ->idle_list is empty regardless of ->nr_idle. After this patch, rebind_workers() no longer performs the nasty idle-rebind retries which require temporary release of gcwq->lock, and both unbinding and rebinding are atomic w.r.t. global_cwq->lock. worker->idle_rebind and global_cwq->rebind_hold are now unnecessary and removed along with the definition of struct idle_rebind. Changed from V1: 1) remove unlikely from too_many_workers(), ->idle_list can be empty anytime, even before this patch, no reason to use unlikely. 2) fix a small rebasing mistake. (which is from rebasing the orignal fixing patch to for-next) 3) add a lot of comments. 4) clear WORKER_REBIND unconditionaly in idle_worker_rebind() tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: WORKER_REBIND is no longer necessary for busy rebinding Because the old unbind/rebinding implementation wasn't atomic w.r.t. GCWQ_DISASSOCIATED manipulation which is protected by global_cwq->lock, we had to use two flags, WORKER_UNBOUND and WORKER_REBIND, to avoid incorrectly losing all NOT_RUNNING bits with back-to-back CPU hotplug operations; otherwise, completion of rebinding while another unbinding is in progress could clear UNBIND prematurely. Now that both unbind/rebinding are atomic w.r.t. GCWQ_DISASSOCIATED, there's no need to use two flags. Just one is enough. Don't use WORKER_REBIND for busy rebinding. tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: WORKER_REBIND is no longer necessary for idle rebinding Now both worker destruction and idle rebinding remove the worker from idle list while it's still idle, so list_empty(&worker->entry) can be used to test whether either is pending and WORKER_DIE to distinguish between the two instead making WORKER_REBIND unnecessary. Use list_empty(&worker->entry) to determine whether destruction or rebinding is pending. This simplifies worker state transitions. WORKER_REBIND is not needed anymore. Remove it. tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: rename manager_mutex to assoc_mutex Now that manager_mutex's role has changed from synchronizing manager role to excluding hotplug against manager, the name is misleading. As it is protecting the CPU-association of the gcwq now, rename it to assoc_mutex. This patch is pure rename and doesn't introduce any functional change. tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use __cpuinit instead of __devinit for cpu callbacks For workqueue hotplug callbacks, it makes less sense to use __devinit which discards the memory after boot if !HOTPLUG. __cpuinit, which discards the memory after boot if !HOTPLUG_CPU fits better. tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix possible stall on try_to_grab_pending() of a delayed work item Currently, when try_to_grab_pending() grabs a delayed work item, it leaves its linked work items alone on the delayed_works. The linked work items are always NO_COLOR and will cause future cwq_activate_first_delayed() increase cwq->nr_active incorrectly, and may cause the whole cwq to stall. For example, state: cwq->max_active = 1, cwq->nr_active = 1 one work in cwq->pool, many in cwq->delayed_works. step1: try_to_grab_pending() removes a work item from delayed_works but leaves its NO_COLOR linked work items on it. step2: Later on, cwq_activate_first_delayed() activates the linked work item increasing ->nr_active. step3: cwq->nr_active = 1, but all activated work items of the cwq are NO_COLOR. When they finish, cwq->nr_active will not be decreased due to NO_COLOR, and no further work items will be activated from cwq->delayed_works. the cwq stalls. Fix it by ensuring the target work item is activated before stealing PENDING in try_to_grab_pending(). This ensures that all the linked work items are activated without incorrectly bumping cwq->nr_active. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@kernel.org workqueue: reimplement work_on_cpu() using system_wq The existing work_on_cpu() implementation is hugely inefficient. It creates a new kthread, execute that single function and then let the kthread die on each invocation. Now that system_wq can handle concurrent executions, there's no advantage of doing this. Reimplement work_on_cpu() using system_wq which makes it simpler and way more efficient. stable: While this isn't a fix in itself, it's needed to fix a workqueue related bug in cpufreq/powernow-k8. AFAICS, this shouldn't break other existing users. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Jiri Kosina <jkosina@suse.cz> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Bjorn Helgaas <bhelgaas@google.com> Cc: Len Brown <lenb@kernel.org> Cc: Rafael J. Wysocki <rjw@sisk.pl> Cc: stable@vger.kernel.org workqueue: introduce cwq_set_max_active() helper for thaw_workqueues() Using a helper instead of open code makes thaw_workqueues() clearer. The helper will also be used by the next patch. tj: Slight update to comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use cwq_set_max_active() helper for workqueue_set_max_active() workqueue_set_max_active() may increase ->max_active without activating delayed works and may make the activation order differ from the queueing order. Both aren't strictly bugs but the resulting behavior could be a bit odd. To make things more consistent, use cwq_set_max_active() helper which immediately makes use of the newly increased max_mactive if there are delayed work items and also keeps the activation order. tj: Slight update to description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: remove spurious WARN_ON_ONCE(in_irq()) from try_to_grab_pending() e0aecdd874 ("workqueue: use irqsafe timer for delayed_work") made try_to_grab_pending() safe to use from irq context but forgot to remove WARN_ON_ONCE(in_irq()). Remove it. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Fengguang Wu <fengguang.wu@intel.com> workqueue: cancel_delayed_work() should return %false if work item is idle 57b30ae77b ("workqueue: reimplement cancel_delayed_work() using try_to_grab_pending()") made cancel_delayed_work() always return %true unless someone else is also trying to cancel the work item, which is broken - if the target work item is idle, the return value should be %false. try_to_grab_pending() indicates that the target work item was idle by zero return value. Use it for return. Note that this brings cancel_delayed_work() in line with __cancel_work_timer() in return value handling. Signed-off-by: Dan Magenheimer <dan.magenheimer@oracle.com> Signed-off-by: Tejun Heo <tj@kernel.org> LKML-Reference: <444a6439-b1a4-4740-9e7e-bc37267cfe73@default> workqueue: exit rescuer_thread() as TASK_RUNNING A rescue thread exiting TASK_INTERRUPTIBLE can lead to a task scheduling off, never to be seen again. In the case where this occurred, an exiting thread hit reiserfs homebrew conditional resched while holding a mutex, bringing the box to its knees. PID: 18105 TASK: ffff8807fd412180 CPU: 5 COMMAND: "kdmflush" #0 [ffff8808157e7670] schedule at ffffffff8143f489 #1 [ffff8808157e77b8] reiserfs_get_block at ffffffffa038ab2d [reiserfs] #2 [ffff8808157e79a8] __block_write_begin at ffffffff8117fb14 #3 [ffff8808157e7a98] reiserfs_write_begin at ffffffffa0388695 [reiserfs] #4 [ffff8808157e7ad8] generic_perform_write at ffffffff810ee9e2 #5 [ffff8808157e7b58] generic_file_buffered_write at ffffffff810eeb41 #6 [ffff8808157e7ba8] __generic_file_aio_write at ffffffff810f1a3a #7 [ffff8808157e7c58] generic_file_aio_write at ffffffff810f1c88 #8 [ffff8808157e7cc8] do_sync_write at ffffffff8114f850 #9 [ffff8808157e7dd8] do_acct_process at ffffffff810a268f [exception RIP: kernel_thread_helper] RIP: ffffffff8144a5c0 RSP: ffff8808157e7f58 RFLAGS: 00000202 RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffffffff8107af60 RDI: ffff8803ee491d18 RBP: 0000000000000000 R8: 0000000000000000 R9: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 Signed-off-by: Mike Galbraith <mgalbraith@suse.de> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org workqueue: mod_delayed_work_on() shouldn't queue timer on 0 delay 8376fe22c7 ("workqueue: implement mod_delayed_work[_on]()") implemented mod_delayed_work[_on]() using the improved try_to_grab_pending(). The function is later used, among others, to replace [__]candel_delayed_work() + queue_delayed_work() combinations. Unfortunately, a delayed_work item w/ zero @delay is handled slightly differently by mod_delayed_work_on() compared to queue_delayed_work_on(). The latter skips timer altogether and directly queues it using queue_work_on() while the former schedules timer which will expire on the closest tick. This means, when @delay is zero, that [__]cancel_delayed_work() + queue_delayed_work_on() makes the target item immediately executable while mod_delayed_work_on() may induce delay of upto a full tick. This somewhat subtle difference breaks some of the converted users. e.g. block queue plugging uses delayed_work for deferred processing and uses mod_delayed_work_on() when the queue needs to be immediately unplugged. The above problem manifested as noticeably higher number of context switches under certain circumstances. The difference in behavior was caused by missing special case handling for 0 delay in mod_delayed_work_on() compared to queue_delayed_work_on(). Joonsoo Kim posted a patch to add it - ("workqueue: optimize mod_delayed_work_on() when @delay == 0")[1]. The patch was queued for 3.8 but it was described as optimization and I missed that it was a correctness issue. As both queue_delayed_work_on() and mod_delayed_work_on() use __queue_delayed_work() for queueing, it seems that the better approach is to move the 0 delay special handling to the function instead of duplicating it in mod_delayed_work_on(). Fix the problem by moving 0 delay special case handling from queue_delayed_work_on() to __queue_delayed_work(). This replaces Joonsoo's patch. [1] http://thread.gmane.org/gmane.linux.kernel/1379011/focus=1379012 Signed-off-by: Tejun Heo <tj@kernel.org> Reported-and-tested-by: Anders Kaseorg <andersk@MIT.EDU> Reported-and-tested-by: Zlatko Calusic <zlatko.calusic@iskon.hr> LKML-Reference: <alpine.DEB.2.00.1211280953350.26602@dr-wily.mit.edu> LKML-Reference: <50A78AA9.5040904@iskon.hr> Cc: Joonsoo Kim <js1304@gmail.com> workqueue: trivial fix for return statement in work_busy() Return type of work_busy() is unsigned int. There is return statement returning boolean value, 'false' in work_busy(). It is not problem, because 'false' may be treated '0'. However, fixing it would make code robust. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: add WARN_ON_ONCE() on CPU number to wq_worker_waking_up() Recently, workqueue code has gone through some changes and we found some bugs related to concurrency management operations happening on the wrong CPU. When a worker is concurrency managed (!WORKER_NOT_RUNNIG), it should be bound to its associated cpu and woken up to that cpu. Add WARN_ON_ONCE() to verify this. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: convert BUG_ON()s in __queue_delayed_work() to WARN_ON_ONCE()s 8852aac25e ("workqueue: mod_delayed_work_on() shouldn't queue timer on 0 delay") unexpectedly uncovered a very nasty abuse of delayed_work in megaraid - it allocated work_struct, casted it to delayed_work and then pass that into queue_delayed_work(). Previously, this was okay because 0 @delay short-circuited to queue_work() before doing anything with delayed_work. 8852aac25e moved 0 @delay test into __queue_delayed_work() after sanity check on delayed_work making megaraid trigger BUG_ON(). Although megaraid is already fixed by c1d390d8e6 ("megaraid: fix BUG_ON() from incorrect use of delayed work"), this patch converts BUG_ON()s in __queue_delayed_work() to WARN_ON_ONCE()s so that such abusers, if there are more, trigger warning but don't crash the machine. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Xiaotian Feng <xtfeng@gmail.com> wq Change-Id: Ia3c507777a995f32bf6b40dc8318203e53134229 Signed-off-by: franciscofranco <franciscofranco.1990@gmail.com>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 21, 2016
This moves ARM over to the asm-generic/unaligned.h header. This has the benefit of better code generated especially for ARMv7 on gcc 4.7+ compilers. As Arnd Bergmann, points out: The asm-generic version uses the "struct" version for native-endian unaligned access and the "byteshift" version for the opposite endianess. The current ARM version however uses the "byteshift" implementation for both. Thanks to Nicolas Pitre for the excellent analysis: Test case: int foo (int *x) { return get_unaligned(x); } long long bar (long long *x) { return get_unaligned(x); } With the current ARM version: foo: ldrb r3, [r0, stratosk#2] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 2B], MEM[(const u8 *)x_1(D) + 2B] ldrb r1, [r0, stratosk#1] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 1B], MEM[(const u8 *)x_1(D) + 1B] ldrb r2, [r0, #0] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D)], MEM[(const u8 *)x_1(D)] mov r3, r3, asl #16 @ tmp154, MEM[(const u8 *)x_1(D) + 2B], ldrb r0, [r0, stratosk#3] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 3B], MEM[(const u8 *)x_1(D) + 3B] orr r3, r3, r1, asl aosp-mirror#8 @, tmp155, tmp154, MEM[(const u8 *)x_1(D) + 1B], orr r3, r3, r2 @ tmp157, tmp155, MEM[(const u8 *)x_1(D)] orr r0, r3, r0, asl #24 @,, tmp157, MEM[(const u8 *)x_1(D) + 3B], bx lr @ bar: stmfd sp!, {r4, r5, r6, r7} @, mov r2, #0 @ tmp184, ldrb r5, [r0, aosp-mirror#6] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 6B], MEM[(const u8 *)x_1(D) + 6B] ldrb r4, [r0, aosp-mirror#5] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 5B], MEM[(const u8 *)x_1(D) + 5B] ldrb ip, [r0, stratosk#2] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 2B], MEM[(const u8 *)x_1(D) + 2B] ldrb r1, [r0, aosp-mirror#4] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 4B], MEM[(const u8 *)x_1(D) + 4B] mov r5, r5, asl #16 @ tmp175, MEM[(const u8 *)x_1(D) + 6B], ldrb r7, [r0, stratosk#1] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 1B], MEM[(const u8 *)x_1(D) + 1B] orr r5, r5, r4, asl aosp-mirror#8 @, tmp176, tmp175, MEM[(const u8 *)x_1(D) + 5B], ldrb r6, [r0, aosp-mirror#7] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 7B], MEM[(const u8 *)x_1(D) + 7B] orr r5, r5, r1 @ tmp178, tmp176, MEM[(const u8 *)x_1(D) + 4B] ldrb r4, [r0, #0] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D)], MEM[(const u8 *)x_1(D)] mov ip, ip, asl #16 @ tmp188, MEM[(const u8 *)x_1(D) + 2B], ldrb r1, [r0, stratosk#3] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 3B], MEM[(const u8 *)x_1(D) + 3B] orr ip, ip, r7, asl aosp-mirror#8 @, tmp189, tmp188, MEM[(const u8 *)x_1(D) + 1B], orr r3, r5, r6, asl #24 @,, tmp178, MEM[(const u8 *)x_1(D) + 7B], orr ip, ip, r4 @ tmp191, tmp189, MEM[(const u8 *)x_1(D)] orr ip, ip, r1, asl #24 @, tmp194, tmp191, MEM[(const u8 *)x_1(D) + 3B], mov r1, r3 @, orr r0, r2, ip @ tmp171, tmp184, tmp194 ldmfd sp!, {r4, r5, r6, r7} bx lr In both cases the code is slightly suboptimal. One may wonder why wasting r2 with the constant 0 in the second case for example. And all the mov's could be folded in subsequent orr's, etc. Now with the asm-generic version: foo: ldr r0, [r0, #0] @ unaligned @,* x bx lr @ bar: mov r3, r0 @ x, x ldr r0, [r0, #0] @ unaligned @,* x ldr r1, [r3, aosp-mirror#4] @ unaligned @, bx lr @ This is way better of course, but only because this was compiled for ARMv7. In this case the compiler knows that the hardware can do unaligned word access. This isn't that obvious for foo(), but if we remove the get_unaligned() from bar as follows: long long bar (long long *x) {return *x; } then the resulting code is: bar: ldmia r0, {r0, r1} @ x,, bx lr @ So this proves that the presumed aligned vs unaligned cases does have influence on the instructions the compiler may use and that the above unaligned code results are not just an accident. Still... this isn't fully conclusive without at least looking at the resulting assembly fron a pre ARMv6 compilation. Let's see with an ARMv5 target: foo: ldrb r3, [r0, #0] @ zero_extendqisi2 @ tmp139,* x ldrb r1, [r0, stratosk#1] @ zero_extendqisi2 @ tmp140, ldrb r2, [r0, stratosk#2] @ zero_extendqisi2 @ tmp143, ldrb r0, [r0, stratosk#3] @ zero_extendqisi2 @ tmp146, orr r3, r3, r1, asl aosp-mirror#8 @, tmp142, tmp139, tmp140, orr r3, r3, r2, asl #16 @, tmp145, tmp142, tmp143, orr r0, r3, r0, asl #24 @,, tmp145, tmp146, bx lr @ bar: stmfd sp!, {r4, r5, r6, r7} @, ldrb r2, [r0, #0] @ zero_extendqisi2 @ tmp139,* x ldrb r7, [r0, stratosk#1] @ zero_extendqisi2 @ tmp140, ldrb r3, [r0, aosp-mirror#4] @ zero_extendqisi2 @ tmp149, ldrb r6, [r0, aosp-mirror#5] @ zero_extendqisi2 @ tmp150, ldrb r5, [r0, stratosk#2] @ zero_extendqisi2 @ tmp143, ldrb r4, [r0, aosp-mirror#6] @ zero_extendqisi2 @ tmp153, ldrb r1, [r0, aosp-mirror#7] @ zero_extendqisi2 @ tmp156, ldrb ip, [r0, stratosk#3] @ zero_extendqisi2 @ tmp146, orr r2, r2, r7, asl aosp-mirror#8 @, tmp142, tmp139, tmp140, orr r3, r3, r6, asl aosp-mirror#8 @, tmp152, tmp149, tmp150, orr r2, r2, r5, asl #16 @, tmp145, tmp142, tmp143, orr r3, r3, r4, asl #16 @, tmp155, tmp152, tmp153, orr r0, r2, ip, asl #24 @,, tmp145, tmp146, orr r1, r3, r1, asl #24 @,, tmp155, tmp156, ldmfd sp!, {r4, r5, r6, r7} bx lr Compared to the initial results, this is really nicely optimized and I couldn't do much better if I were to hand code it myself. Signed-off-by: Rob Herring <rob.herring@calxeda.com> Reviewed-by: Nicolas Pitre <nico@linaro.org> Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> modified for Mako from kernel.org reference Signed-off-by: faux123 <reioux@gmail.com>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 21, 2016
workqueue: change BUG_ON() to WARN_ON() This BUG_ON() can be triggered if you call schedule_work() before calling INIT_WORK(). It is a bug definitely, but it's nicer to just print a stack trace and return. Reported-by: Matt Renzelmann <mjr@cs.wisc.edu> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: Catch more locking problems with flush_work() If a workqueue is flushed with flush_work() lockdep checking can be circumvented. For example: static DEFINE_MUTEX(mutex); static void my_work(struct work_struct *w) { mutex_lock(&mutex); mutex_unlock(&mutex); } static DECLARE_WORK(work, my_work); static int __init start_test_module(void) { schedule_work(&work); return 0; } module_init(start_test_module); static void __exit stop_test_module(void) { mutex_lock(&mutex); flush_work(&work); mutex_unlock(&mutex); } module_exit(stop_test_module); would not always print a warning when flush_work() was called. In this trivial example nothing could go wrong since we are guaranteed module_init() and module_exit() don't run concurrently, but if the work item is schedule asynchronously we could have a scenario where the work item is running just at the time flush_work() is called resulting in a classic ABBA locking problem. Add a lockdep hint by acquiring and releasing the work item lockdep_map in flush_work() so that we always catch this potential deadlock scenario. Signed-off-by: Stephen Boyd <sboyd@codeaurora.org> Reviewed-by: Yong Zhang <yong.zhang0@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> lockdep: fix oops in processing workqueue Under memory load, on x86_64, with lockdep enabled, the workqueue's process_one_work() has been seen to oops in __lock_acquire(), barfing on a 0xffffffff00000000 pointer in the lockdep_map's class_cache[]. Because it's permissible to free a work_struct from its callout function, the map used is an onstack copy of the map given in the work_struct: and that copy is made without any locking. Surprisingly, gcc (4.5.1 in Hugh's case) uses "rep movsl" rather than "rep movsq" for that structure copy: which might race with a workqueue user's wait_on_work() doing lock_map_acquire() on the source of the copy, putting a pointer into the class_cache[], but only in time for the top half of that pointer to be copied to the destination map. Boom when process_one_work() subsequently does lock_map_acquire() on its onstack copy of the lockdep_map. Fix this, and a similar instance in call_timer_fn(), with a lockdep_copy_map() function which additionally NULLs the class_cache[]. Note: this oops was actually seen on 3.4-next, where flush_work() newly does the racing lock_map_acquire(); but Tejun points out that 3.4 and earlier are already vulnerable to the same through wait_on_work(). * Patch orginally from Peter. Hugh modified it a bit and wrote the description. Signed-off-by: Peter Zijlstra <peterz@infradead.org> Reported-by: Hugh Dickins <hughd@google.com> LKML-Reference: <alpine.LSU.2.00.1205070951170.1544@eggly.anvils> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: perform cpu down operations from low priority cpu_notifier() Currently, all workqueue cpu hotplug operations run off CPU_PRI_WORKQUEUE which is higher than normal notifiers. This is to ensure that workqueue is up and running while bringing up a CPU before other notifiers try to use workqueue on the CPU. Per-cpu workqueues are supposed to remain working and bound to the CPU for normal CPU_DOWN_PREPARE notifiers. This holds mostly true even with workqueue offlining running with higher priority because workqueue CPU_DOWN_PREPARE only creates a bound trustee thread which runs the per-cpu workqueue without concurrency management without explicitly detaching the existing workers. However, if the trustee needs to create new workers, it creates unbound workers which may wander off to other CPUs while CPU_DOWN_PREPARE notifiers are in progress. Furthermore, if the CPU down is cancelled, the per-CPU workqueue may end up with workers which aren't bound to the CPU. While reliably reproducible with a convoluted artificial test-case involving scheduling and flushing CPU burning work items from CPU down notifiers, this isn't very likely to happen in the wild, and, even when it happens, the effects are likely to be hidden by the following successful CPU down. Fix it by using different priorities for up and down notifiers - high priority for up operations and low priority for down operations. Workqueue cpu hotplug operations will soon go through further cleanup. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: drop CPU_DYING notifier operation Workqueue used CPU_DYING notification to mark GCWQ_DISASSOCIATED. This was necessary because workqueue's CPU_DOWN_PREPARE happened before other DOWN_PREPARE notifiers and workqueue needed to stay associated across the rest of DOWN_PREPARE. After the previous patch, workqueue's DOWN_PREPARE happens after others and can set GCWQ_DISASSOCIATED directly. Drop CPU_DYING and let the trustee set GCWQ_DISASSOCIATED after disabling concurrency management. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: ROGUE workers are UNBOUND workers Currently, WORKER_UNBOUND is used to mark workers for the unbound global_cwq and WORKER_ROGUE is used to mark workers for disassociated per-cpu global_cwqs. Both are used to make the marked worker skip concurrency management and the only place they make any difference is in worker_enter_idle() where WORKER_ROGUE is used to skip scheduling idle timer, which can easily be replaced with trustee state testing. This patch replaces WORKER_ROGUE with WORKER_UNBOUND and drops WORKER_ROGUE. This is to prepare for removing trustee and handling disassociated global_cwqs as unbound. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: use mutex for global_cwq manager exclusion POOL_MANAGING_WORKERS is used to ensure that at most one worker takes the manager role at any given time on a given global_cwq. Trustee later hitched on it to assume manager adding blocking wait for the bit. As trustee already needed a custom wait mechanism, waiting for MANAGING_WORKERS was rolled into the same mechanism. Trustee is scheduled to be removed. This patch separates out MANAGING_WORKERS wait into per-pool mutex. Workers use mutex_trylock() to test for manager role and trustee uses mutex_lock() to claim manager roles. gcwq_claim/release_management() helpers are added to grab and release manager roles of all pools on a global_cwq. gcwq_claim_management() always grabs pool manager mutexes in ascending pool index order and uses pool index as lockdep subclass. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: drop @bind from create_worker() Currently, create_worker()'s callers are responsible for deciding whether the newly created worker should be bound to the associated CPU and create_worker() sets WORKER_UNBOUND only for the workers for the unbound global_cwq. Creation during normal operation is always via maybe_create_worker() and @bind is true. For workers created during hotplug, @bind is false. Normal operation path is planned to be used even while the CPU is going through hotplug operations or offline and this static decision won't work. Drop @bind from create_worker() and decide whether to bind by looking at GCWQ_DISASSOCIATED. create_worker() will also set WORKER_UNBOUND autmatically if disassociated. To avoid flipping GCWQ_DISASSOCIATED while create_worker() is in progress, the flag is now allowed to be changed only while holding all manager_mutexes on the global_cwq. This requires that GCWQ_DISASSOCIATED is not cleared behind trustee's back. CPU_ONLINE no longer clears DISASSOCIATED before flushing trustee, which clears DISASSOCIATED before rebinding remaining workers if asked to release. For cases where trustee isn't around, CPU_ONLINE clears DISASSOCIATED after flushing trustee. Also, now, first_idle has UNBOUND set on creation which is explicitly cleared by CPU_ONLINE while binding it. These convolutions will soon be removed by further simplification of CPU hotplug path. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: reimplement CPU online rebinding to handle idle workers Currently, if there are left workers when a CPU is being brough back online, the trustee kills all idle workers and scheduled rebind_work so that they re-bind to the CPU after the currently executing work is finished. This works for busy workers because concurrency management doesn't try to wake up them from scheduler callbacks, which require the target task to be on the local run queue. The busy worker bumps concurrency counter appropriately as it clears WORKER_UNBOUND from the rebind work item and it's bound to the CPU before returning to the idle state. To reduce CPU on/offlining overhead (as many embedded systems use it for powersaving) and simplify the code path, workqueue is planned to be modified to retain idle workers across CPU on/offlining. This patch reimplements CPU online rebinding such that it can also handle idle workers. As noted earlier, due to the local wakeup requirement, rebinding idle workers is tricky. All idle workers must be re-bound before scheduler callbacks are enabled. This is achieved by interlocking idle re-binding. Idle workers are requested to re-bind and then hold until all idle re-binding is complete so that no bound worker starts executing work item. Only after all idle workers are re-bound and parked, CPU_ONLINE proceeds to release them and queue rebind work item to busy workers thus guaranteeing scheduler callbacks aren't invoked until all idle workers are ready. worker_rebind_fn() is renamed to busy_worker_rebind_fn() and idle_worker_rebind() for idle workers is added. Rebinding logic is moved to rebind_workers() and now called from CPU_ONLINE after flushing trustee. While at it, add CPU sanity check in worker_thread(). Note that now a worker may become idle or the manager between trustee release and rebinding during CPU_ONLINE. As the previous patch updated create_worker() so that it can be used by regular manager while unbound and this patch implements idle re-binding, this is safe. This prepares for removal of trustee and keeping idle workers across CPU hotplugs. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: don't butcher idle workers on an offline CPU Currently, during CPU offlining, after all pending work items are drained, the trustee butchers all workers. Also, on CPU onlining failure, workqueue_cpu_callback() ensures that the first idle worker is destroyed. Combined, these guarantee that an offline CPU doesn't have any worker for it once all the lingering work items are finished. This guarantee isn't really necessary and makes CPU on/offlining more expensive than needs to be, especially for platforms which use CPU hotplug for powersaving. This patch lets offline CPUs removes idle worker butchering from the trustee and let a CPU which failed onlining keep the created first worker. The first worker is created if the CPU doesn't have any during CPU_DOWN_PREPARE and started right away. If onlining succeeds, the rebind_workers() call in CPU_ONLINE will rebind it like any other workers. If onlining fails, the worker is left alone till the next try. This makes CPU hotplugs cheaper by allowing global_cwqs to keep workers across them and simplifies code. Note that trustee doesn't re-arm idle timer when it's done and thus the disassociated global_cwq will keep all workers until it comes back online. This will be improved by further patches. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: remove CPU offline trustee With the previous changes, a disassociated global_cwq now can run as an unbound one on its own - it can create workers as necessary to drain remaining works after the CPU has been brought down and manage the number of workers using the usual idle timer mechanism making trustee completely redundant except for the actual unbinding operation. This patch removes the trustee and let a disassociated global_cwq manage itself. Unbinding is moved to a work item (for CPU affinity) which is scheduled and flushed from CPU_DONW_PREPARE. This patch moves nr_running clearing outside gcwq and manager locks to simplify the code. As nr_running is unused at the point, this is safe. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: simplify CPU hotplug code With trustee gone, CPU hotplug code can be simplified. * gcwq_claim/release_management() now grab and release gcwq lock too respectively and gained _and_lock and _and_unlock postfixes. * All CPU hotplug logic was implemented in workqueue_cpu_callback() which was called by workqueue_cpu_up/down_callback() for the correct priority. This was because up and down paths shared a lot of logic, which is no longer true. Remove workqueue_cpu_callback() and move all hotplug logic into the two actual callbacks. This patch doesn't make any functional changes. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl> workqueue: fix spurious CPU locality WARN from process_one_work() 25511a4776 "workqueue: reimplement CPU online rebinding to handle idle workers" added CPU locality sanity check in process_one_work(). It triggers if a worker is executing on a different CPU without UNBOUND or REBIND set. This works for all normal workers but rescuers can trigger this spuriously when they're serving the unbound or a disassociated global_cwq - rescuers don't have either flag set and thus its gcwq->cpu can be a different value including %WORK_CPU_UNBOUND. Fix it by additionally testing %GCWQ_DISASSOCIATED. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> LKML-Refence: <20120721213656.GA7783@linux.vnet.ibm.com> workqueue: reorder queueing functions so that _on() variants are on top Currently, queue/schedule[_delayed]_work_on() are located below the counterpart without the _on postifx even though the latter is usually implemented using the former. Swap them. This is cleanup and doesn't cause any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: make queueing functions return bool All queueing functions return 1 on success, 0 if the work item was already pending. Update them to return bool instead. This signifies better that they don't return 0 / -errno. This is cleanup and doesn't cause any functional difference. While at it, fix comment opening for schedule_work_on(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: add missing smp_wmb() in process_one_work() WORK_STRUCT_PENDING is used to claim ownership of a work item and process_one_work() releases it before starting execution. When someone else grabs PENDING, all pre-release updates to the work item should be visible and all updates made by the new owner should happen afterwards. Grabbing PENDING uses test_and_set_bit() and thus has a full barrier; however, clearing doesn't have a matching wmb. Given the preceding spin_unlock and use of clear_bit, I don't believe this can be a problem on an actual machine and there hasn't been any related report but it still is theretically possible for clear_pending to permeate upwards and happen before work->entry update. Add an explicit smp_wmb() before work_clear_pending(). Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: stable@vger.kernel.org workqueue: disable irq while manipulating PENDING Queueing operations use WORK_STRUCT_PENDING_BIT to synchronize access to the target work item. They first try to claim the bit and proceed with queueing only after that succeeds and there's a window between PENDING being set and the actual queueing where the task can be interrupted or preempted. There's also a similar window in process_one_work() when clearing PENDING. A work item is dequeued, gcwq->lock is released and then PENDING is cleared and the worker might get interrupted or preempted between releasing gcwq->lock and clearing PENDING. cancel[_delayed]_work_sync() tries to claim or steal PENDING. The function assumes that a work item with PENDING is either queued or in the process of being [de]queued. In the latter case, it busy-loops until either the work item loses PENDING or is queued. If canceling coincides with the above described interrupts or preemptions, the canceling task will busy-loop while the queueing or executing task is preempted. This patch keeps irq disabled across claiming PENDING and actual queueing and moves PENDING clearing in process_one_work() inside gcwq->lock so that busy looping from PENDING && !queued doesn't wait for interrupted/preempted tasks. Note that, in process_one_work(), setting last CPU and clearing PENDING got merged into single operation. This removes possible long busy-loops and will allow using try_to_grab_pending() from bh and irq contexts. v2: __queue_work() was testing preempt_count() to ensure that the caller has disabled preemption. This triggers spuriously if !CONFIG_PREEMPT_COUNT. Use preemptible() instead. Reported by Fengguang Wu. v3: Disable irq instead of preemption. IRQ will be disabled while grabbing gcwq->lock later anyway and this allows using try_to_grab_pending() from bh and irq contexts. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Fengguang Wu <fengguang.wu@intel.com> workqueue: set delayed_work->timer function on initialization delayed_work->timer.function is currently initialized during queue_delayed_work_on(). Export delayed_work_timer_fn() and set delayed_work timer function during delayed_work initialization together with other fields. This ensures the timer function is always valid on an initialized delayed_work. This is to help mod_delayed_work() implementation. To detect delayed_work users which diddle with the internal timer, trigger WARN if timer function doesn't match on queue. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: unify local CPU queueing handling Queueing functions have been using different methods to determine the local CPU. * queue_work() superflously uses get/put_cpu() to acquire and hold the local CPU across queue_work_on(). * delayed_work_timer_fn() uses smp_processor_id(). * queue_delayed_work() calls queue_delayed_work_on() with -1 @cpu which is interpreted as the local CPU. * flush_delayed_work[_sync]() were using raw_smp_processor_id(). * __queue_work() interprets %WORK_CPU_UNBOUND as local CPU if the target workqueue is bound one but nobody uses this. This patch converts all functions to uniformly use %WORK_CPU_UNBOUND to indicate local CPU and use the local binding feature of __queue_work(). unlikely() is dropped from %WORK_CPU_UNBOUND handling in __queue_work(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix zero @delay handling of queue_delayed_work_on() If @delay is zero and the dealyed_work is idle, queue_delayed_work() queues it for immediate execution; however, queue_delayed_work_on() lacks this logic and always goes through timer regardless of @delay. This patch moves 0 @delay handling logic from queue_delayed_work() to queue_delayed_work_on() so that both functions behave the same. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: move try_to_grab_pending() upwards try_to_grab_pending() will be used by to-be-implemented mod_delayed_work[_on](). Move try_to_grab_pending() and related functions above queueing functions. This patch only moves functions around. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: introduce WORK_OFFQ_FLAG_* Low WORK_STRUCT_FLAG_BITS bits of work_struct->data contain WORK_STRUCT_FLAG_* and flush color. If the work item is queued, the rest point to the cpu_workqueue with WORK_STRUCT_CWQ set; otherwise, WORK_STRUCT_CWQ is clear and the bits contain the last CPU number - either a real CPU number or one of WORK_CPU_*. Scheduled addition of mod_delayed_work[_on]() requires an additional flag, which is used only while a work item is off queue. There are more than enough bits to represent off-queue CPU number on both 32 and 64bits. This patch introduces WORK_OFFQ_FLAG_* which occupy the lower part of the @work->data high bits while off queue. This patch doesn't define any actual OFFQ flag yet. Off-queue CPU number is now shifted by WORK_OFFQ_CPU_SHIFT, which adds the number of bits used by OFFQ flags to WORK_STRUCT_FLAG_SHIFT, to make room for OFFQ flags. To avoid shift width warning with large WORK_OFFQ_FLAG_BITS, ulong cast is added to WORK_STRUCT_NO_CPU and, just in case, BUILD_BUG_ON() to check that there are enough bits to accomodate off-queue CPU number is added. This patch doesn't make any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: factor out __queue_delayed_work() from queue_delayed_work_on() This is to prepare for mod_delayed_work[_on]() and doesn't cause any functional difference. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reorganize try_to_grab_pending() and __cancel_timer_work() * Use bool @is_dwork instead of @timer and let try_to_grab_pending() use to_delayed_work() to determine the delayed_work address. * Move timer handling from __cancel_work_timer() to try_to_grab_pending(). * Make try_to_grab_pending() use -EAGAIN instead of -1 for busy-looping and drop the ret local variable. * Add proper function comment to try_to_grab_pending(). This makes the code a bit easier to understand and will ease further changes. This patch doesn't make any functional change. v2: Use @is_dwork instead of @timer. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: mark a work item being canceled as such There can be two reasons try_to_grab_pending() can fail with -EAGAIN. One is when someone else is queueing or deqeueing the work item. With the previous patches, it is guaranteed that PENDING and queued state will soon agree making it safe to busy-retry in this case. The other is if multiple __cancel_work_timer() invocations are racing one another. __cancel_work_timer() grabs PENDING and then waits for running instances of the target work item on all CPUs while holding PENDING and !queued. try_to_grab_pending() invoked from another task will keep returning -EAGAIN while the current owner is waiting. Not distinguishing the two cases is okay because __cancel_work_timer() is the only user of try_to_grab_pending() and it invokes wait_on_work() whenever grabbing fails. For the first case, busy looping should be fine but wait_on_work() doesn't cause any critical problem. For the latter case, the new contender usually waits for the same condition as the current owner, so no unnecessarily extended busy-looping happens. Combined, these make __cancel_work_timer() technically correct even without irq protection while grabbing PENDING or distinguishing the two different cases. While the current code is technically correct, not distinguishing the two cases makes it difficult to use try_to_grab_pending() for other purposes than canceling because it's impossible to tell whether it's safe to busy-retry grabbing. This patch adds a mechanism to mark a work item being canceled. try_to_grab_pending() now disables irq on success and returns -EAGAIN to indicate that grabbing failed but PENDING and queued states are gonna agree soon and it's safe to busy-loop. It returns -ENOENT if the work item is being canceled and it may stay PENDING && !queued for arbitrary amount of time. __cancel_work_timer() is modified to mark the work canceling with WORK_OFFQ_CANCELING after grabbing PENDING, thus making try_to_grab_pending() fail with -ENOENT instead of -EAGAIN. Also, it invokes wait_on_work() iff grabbing failed with -ENOENT. This isn't necessary for correctness but makes it consistent with other future users of try_to_grab_pending(). v2: try_to_grab_pending() was testing preempt_count() to ensure that the caller has disabled preemption. This triggers spuriously if !CONFIG_PREEMPT_COUNT. Use preemptible() instead. Reported by Fengguang Wu. v3: Updated so that try_to_grab_pending() disables irq on success rather than requiring preemption disabled by the caller. This makes busy-looping easier and will allow try_to_grap_pending() to be used from bh/irq contexts. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Fengguang Wu <fengguang.wu@intel.com> workqueue: implement mod_delayed_work[_on]() Workqueue was lacking a mechanism to modify the timeout of an already pending delayed_work. delayed_work users have been working around this using several methods - using an explicit timer + work item, messing directly with delayed_work->timer, and canceling before re-queueing, all of which are error-prone and/or ugly. This patch implements mod_delayed_work[_on]() which behaves similarly to mod_timer() - if the delayed_work is idle, it's queued with the given delay; otherwise, its timeout is modified to the new value. Zero @delay guarantees immediate execution. v2: Updated to reflect try_to_grab_pending() changes. Now safe to be called from bh context. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> workqueue: fix CPU binding of flush_delayed_work[_sync]() delayed_work encodes the workqueue to use and the last CPU in delayed_work->work.data while it's on timer. The target CPU is implicitly recorded as the CPU the timer is queued on and delayed_work_timer_fn() queues delayed_work->work to the CPU it is running on. Unfortunately, this leaves flush_delayed_work[_sync]() no way to find out which CPU the delayed_work was queued for when they try to re-queue after killing the timer. Currently, it chooses the local CPU flush is running on. This can unexpectedly move a delayed_work queued on a specific CPU to another CPU and lead to subtle errors. There isn't much point in trying to save several bytes in struct delayed_work, which is already close to a hundred bytes on 64bit with all debug options turned off. This patch adds delayed_work->cpu to remember the CPU it's queued for. Note that if the timer is migrated during CPU down, the work item could be queued to the downed global_cwq after this change. As a detached global_cwq behaves like an unbound one, this doesn't change much for the delayed_work. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> workqueue: add missing wmb() in clear_work_data() Any operation which clears PENDING should be preceded by a wmb to guarantee that the next PENDING owner sees all the changes made before PENDING release. There are only two places where PENDING is cleared - set_work_cpu_and_clear_pending() and clear_work_data(). The caller of the former already does smp_wmb() but the latter doesn't have any. Move the wmb above set_work_cpu_and_clear_pending() into it and add one to clear_work_data(). There hasn't been any report related to this issue, and, given how clear_work_data() is used, it is extremely unlikely to have caused any actual problems on any architecture. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> workqueue: use enum value to set array size of pools in gcwq Commit 3270476a6c0ce322354df8679652f060d66526dc ('workqueue: reimplement WQ_HIGHPRI using a separate worker_pool') introduce separate worker_pool for HIGHPRI. Although there is NR_WORKER_POOLS enum value which represent size of pools, definition of worker_pool in gcwq doesn't use it. Using it makes code robust and prevent future mistakes. So change code to use this enum value. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: correct req_cpu in trace_workqueue_queue_work() When we do tracing workqueue_queue_work(), it records requested cpu. But, if !(@wq->flag & WQ_UNBOUND) and @cpu is WORK_CPU_UNBOUND, requested cpu is changed as local cpu. In case of @wq->flag & WQ_UNBOUND, above change is not occured, therefore it is reasonable to correct it. Use temporary local variable for storing requested cpu. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: change value of lcpu in __queue_delayed_work_on() We assign cpu id into work struct's data field in __queue_delayed_work_on(). In current implementation, when work is come in first time, current running cpu id is assigned. If we do __queue_delayed_work_on() with CPU A on CPU B, __queue_work() invoked in delayed_work_timer_fn() go into the following sub-optimal path in case of WQ_NON_REENTRANT. gcwq = get_gcwq(cpu); if (wq->flags & WQ_NON_REENTRANT && (last_gcwq = get_work_gcwq(work)) && last_gcwq != gcwq) { Change lcpu to @cpu and rechange lcpu to local cpu if lcpu is WORK_CPU_UNBOUND. It is sufficient to prevent to go into sub-optimal path. tj: Slightly rephrased the comment. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: introduce system_highpri_wq Commit 3270476a6c0ce322354df8679652f060d66526dc ('workqueue: reimplement WQ_HIGHPRI using a separate worker_pool') introduce separate worker pool for HIGHPRI. When we handle busyworkers for gcwq, it can be normal worker or highpri worker. But, we don't consider this difference in rebind_workers(), we use just system_wq for highpri worker. It makes mismatch between cwq->pool and worker->pool. It doesn't make error in current implementation, but possible in the future. Now, we introduce system_highpri_wq to use proper cwq for highpri workers in rebind_workers(). Following patch fix this issue properly. tj: Even apart from rebinding, having system_highpri_wq generally makes sense. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use system_highpri_wq for highpri workers in rebind_workers() In rebind_workers(), we do inserting a work to rebind to cpu for busy workers. Currently, in this case, we use only system_wq. This makes a possible error situation as there is mismatch between cwq->pool and worker->pool. To prevent this, we should use system_highpri_wq for highpri worker to match theses. This implements it. tj: Rephrased comment a bit. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use system_highpri_wq for unbind_work To speed cpu down processing up, use system_highpri_wq. As scheduling priority of workers on it is higher than system_wq and it is not contended by other normal works on this cpu, work on it is processed faster than system_wq. tj: CPU up/downs care quite a bit about latency these days. This shouldn't hurt anything and makes sense. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix checkpatch issues Fixed some checkpatch warnings. tj: adapted to wq/for-3.7 and massaged pr_xxx() format strings a bit. Signed-off-by: Valentin Ilie <valentin.ilie@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> LKML-Reference: <1345326762-21747-1-git-send-email-valentin.ilie@gmail.com> workqueue: make all workqueues non-reentrant By default, each per-cpu part of a bound workqueue operates separately and a work item may be executing concurrently on different CPUs. The behavior avoids some cross-cpu traffic but leads to subtle weirdities and not-so-subtle contortions in the API. * There's no sane usefulness in allowing a single work item to be executed concurrently on multiple CPUs. People just get the behavior unintentionally and get surprised after learning about it. Most either explicitly synchronize or use non-reentrant/ordered workqueue but this is error-prone. * flush_work() can't wait for multiple instances of the same work item on different CPUs. If a work item is executing on cpu0 and then queued on cpu1, flush_work() can only wait for the one on cpu1. Unfortunately, work items can easily cross CPU boundaries unintentionally when the queueing thread gets migrated. This means that if multiple queuers compete, flush_work() can't even guarantee that the instance queued right before it is finished before returning. * flush_work_sync() was added to work around some of the deficiencies of flush_work(). In addition to the usual flushing, it ensures that all currently executing instances are finished before returning. This operation is expensive as it has to walk all CPUs and at the same time fails to address competing queuer case. Incorrectly using flush_work() when flush_work_sync() is necessary is an easy error to make and can lead to bugs which are difficult to reproduce. * Similar problems exist for flush_delayed_work[_sync](). Other than the cross-cpu access concern, there's no benefit in allowing parallel execution and it's plain silly to have this level of contortion for workqueue which is widely used from core code to extremely obscure drivers. This patch makes all workqueues non-reentrant. If a work item is executing on a different CPU when queueing is requested, it is always queued to that CPU. This guarantees that any given work item can be executing on one CPU at maximum and if a work item is queued and executing, both are on the same CPU. The only behavior change which may affect workqueue users negatively is that non-reentrancy overrides the affinity specified by queue_work_on(). On a reentrant workqueue, the affinity specified by queue_work_on() is always followed. Now, if the work item is executing on one of the CPUs, the work item will be queued there regardless of the requested affinity. I've reviewed all workqueue users which request explicit affinity, and, fortunately, none seems to be crazy enough to exploit parallel execution of the same work item. This adds an additional busy_hash lookup if the work item was previously queued on a different CPU. This shouldn't be noticeable under any sane workload. Work item queueing isn't a very high-frequency operation and they don't jump across CPUs all the time. In a micro benchmark to exaggerate this difference - measuring the time it takes for two work items to repeatedly jump between two CPUs a number (10M) of times with busy_hash table densely populated, the difference was around 3%. While the overhead is measureable, it is only visible in pathological cases and the difference isn't huge. This change brings much needed sanity to workqueue and makes its behavior consistent with timer. I think this is the right tradeoff to make. This enables significant simplification of workqueue API. Simplification patches will follow. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: gut flush[_delayed]_work_sync() Now that all workqueues are non-reentrant, flush[_delayed]_work_sync() are equivalent to flush[_delayed]_work(). Drop the separate implementation and make them thin wrappers around flush[_delayed]_work(). * start_flush_work() no longer takes @wait_executing as the only left user - flush_work() - always sets it to %true. * __cancel_work_timer() uses flush_work() instead of wait_on_work(). Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: gut system_nrt[_freezable]_wq() Now that all workqueues are non-reentrant, system[_freezable]_wq() are equivalent to system_nrt[_freezable]_wq(). Replace the latter with wrappers around system[_freezable]_wq(). The wrapping goes through inline functions so that __deprecated can be added easily. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: cosmetic whitespace updates for macro definitions Consistently use the last tab position for '\' line continuation in complex macro definitions. This is to help the following patches. This patch is cosmetic. Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use hotcpu_notifier() for workqueue_cpu_down_callback() workqueue_cpu_down_callback() is used only if HOTPLUG_CPU=y, so hotcpu_notifier() fits better than cpu_notifier(). When HOTPLUG_CPU=y, hotcpu_notifier() and cpu_notifier() are the same. When HOTPLUG_CPU=n, if we use cpu_notifier(), workqueue_cpu_down_callback() will be called during boot to do nothing, and the memory of workqueue_cpu_down_callback() and gcwq_unbind_fn() will be discarded after boot. If we use hotcpu_notifier(), we can avoid the no-op call of workqueue_cpu_down_callback() and the memory of workqueue_cpu_down_callback() and gcwq_unbind_fn() will be discard at build time: $ ls -l kernel/workqueue.o.cpu_notifier kernel/workqueue.o.hotcpu_notifier -rw-rw-r-- 1 laijs laijs 484080 Sep 15 11:31 kernel/workqueue.o.cpu_notifier -rw-rw-r-- 1 laijs laijs 478240 Sep 15 11:31 kernel/workqueue.o.hotcpu_notifier $ size kernel/workqueue.o.cpu_notifier kernel/workqueue.o.hotcpu_notifier text data bss dec hex filename 18513 2387 1221 22121 5669 kernel/workqueue.o.cpu_notifier 18082 2355 1221 21658 549a kernel/workqueue.o.hotcpu_notifier tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reimplement cancel_delayed_work() using try_to_grab_pending() cancel_delayed_work() can't be called from IRQ handlers due to its use of del_timer_sync() and can't cancel work items which are already transferred from timer to worklist. Also, unlike other flush and cancel functions, a canceled delayed_work would still point to the last associated cpu_workqueue. If the workqueue is destroyed afterwards and the work item is re-used on a different workqueue, the queueing code can oops trying to dereference already freed cpu_workqueue. This patch reimplements cancel_delayed_work() using try_to_grab_pending() and set_work_cpu_and_clear_pending(). This allows the function to be called from IRQ handlers and makes its behavior consistent with other flush / cancel functions. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> workqueue: UNBOUND -> REBIND morphing in rebind_workers() should be atomic The compiler may compile the following code into TWO write/modify instructions. worker->flags &= ~WORKER_UNBOUND; worker->flags |= WORKER_REBIND; so the other CPU may temporarily see worker->flags which doesn't have either WORKER_UNBOUND or WORKER_REBIND set and perform local wakeup prematurely. Fix it by using single explicit assignment via ACCESS_ONCE(). Because idle workers have another WORKER_NOT_RUNNING flag, this bug doesn't exist for them; however, update it to use the same pattern for consistency. tj: Applied the change to idle workers too and updated comments and patch description a bit. Change-Id: I9b95f51d146c40c31ba028668d6f412bd74c6026 Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org workqueue: move WORKER_REBIND clearing in rebind_workers() to the end of the function This doesn't make any functional difference and is purely to help the next patch to be simpler. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> workqueue: fix possible deadlock in idle worker rebinding Currently, rebind_workers() and idle_worker_rebind() are two-way interlocked. rebind_workers() waits for idle workers to finish rebinding and rebound idle workers wait for rebind_workers() to finish rebinding busy workers before proceeding. Unfortunately, this isn't enough. The second wait from idle workers is implemented as follows. wait_event(gcwq->rebind_hold, !(worker->flags & WORKER_REBIND)); rebind_workers() clears WORKER_REBIND, wakes up the idle workers and then returns. If CPU hotplug cycle happens again before one of the idle workers finishes the above wait_event(), rebind_workers() will repeat the first part of the handshake - set WORKER_REBIND again and wait for the idle worker to finish rebinding - and this leads to deadlock because the idle worker would be waiting for WORKER_REBIND to clear. This is fixed by adding another interlocking step at the end - rebind_workers() now waits for all the idle workers to finish the above WORKER_REBIND wait before returning. This ensures that all rebinding steps are complete on all idle workers before the next hotplug cycle can happen. This problem was diagnosed by Lai Jiangshan who also posted a patch to fix the issue, upon which this patch is based. This is the minimal fix and further patches are scheduled for the next merge window to simplify the CPU hotplug path. Signed-off-by: Tejun Heo <tj@kernel.org> Original-patch-by: Lai Jiangshan <laijs@cn.fujitsu.com> LKML-Reference: <1346516916-1991-3-git-send-email-laijs@cn.fujitsu.com> workqueue: restore POOL_MANAGING_WORKERS This patch restores POOL_MANAGING_WORKERS which was replaced by pool->manager_mutex by 6037315269 "workqueue: use mutex for global_cwq manager exclusion". There's a subtle idle worker depletion bug across CPU hotplug events and we need to distinguish an actual manager and CPU hotplug preventing management. POOL_MANAGING_WORKERS will be used for the former and manager_mutex the later. This patch just lays POOL_MANAGING_WORKERS on top of the existing manager_mutex and doesn't introduce any synchronization changes. The next patch will update it. Note that this patch fixes a non-critical anomaly where too_many_workers() may return %true spuriously while CPU hotplug is in progress. While the issue could schedule idle timer spuriously, it didn't trigger any actual misbehavior. tj: Rewrote patch description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix possible idle worker depletion across CPU hotplug To simplify both normal and CPU hotplug paths, worker management is prevented while CPU hoplug is in progress. This is achieved by CPU hotplug holding the same exclusion mechanism used by workers to ensure there's only one manager per pool. If someone else seems to be performing the manager role, workers proceed to execute work items. CPU hotplug using the same mechanism can lead to idle worker depletion because all workers could proceed to execute work items while CPU hotplug is in progress and CPU hotplug itself wouldn't actually perform the worker management duty - it doesn't guarantee that there's an idle worker left when it releases management. This idle worker depletion, under extreme circumstances, can break forward-progress guarantee and thus lead to deadlock. This patch fixes the bug by using separate mechanisms for manager exclusion among workers and hotplug exclusion. For manager exclusion, POOL_MANAGING_WORKERS which was restored by the previous patch is used. pool->manager_mutex is now only used for exclusion between the elected manager and CPU hotplug. The elected manager won't proceed without holding pool->manager_mutex. This ensures that the worker which won the manager position can't skip managing while CPU hotplug is in progress. It will block on manager_mutex and perform management after CPU hotplug is complete. Note that hotplug may happen while waiting for manager_mutex. A manager isn't either on idle or busy list and thus the hoplug code can't unbind/rebind it. Make the manager handle its own un/rebinding. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: always clear WORKER_REBIND in busy_worker_rebind_fn() busy_worker_rebind_fn() didn't clear WORKER_REBIND if rebinding failed (CPU is down again). This used to be okay because the flag wasn't used for anything else. However, after 25511a477 "workqueue: reimplement CPU online rebinding to handle idle workers", WORKER_REBIND is also used to command idle workers to rebind. If not cleared, the worker may confuse the next CPU_UP cycle by having REBIND spuriously set or oops / get stuck by prematurely calling idle_worker_rebind(). WARNING: at /work/os/wq/kernel/workqueue.c:1323 worker_thread+0x4cd/0x5 00() Hardware name: Bochs Modules linked in: test_wq(O-) Pid: 33, comm: kworker/1:1 Tainted: G O 3.6.0-rc1-work+ #3 Call Trace: [<ffffffff8109039f>] warn_slowpath_common+0x7f/0xc0 [<ffffffff810903fa>] warn_slowpath_null+0x1a/0x20 [<ffffffff810b3f1d>] worker_thread+0x4cd/0x500 [<ffffffff810bc16e>] kthread+0xbe/0xd0 [<ffffffff81bd2664>] kernel_thread_helper+0x4/0x10 ---[ end trace e977cf20f4661968 ]--- BUG: unable to handle kernel NULL pointer dereference at (null) IP: [<ffffffff810b3db0>] worker_thread+0x360/0x500 PGD 0 Oops: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC Modules linked in: test_wq(O-) CPU 0 Pid: 33, comm: kworker/1:1 Tainted: G W O 3.6.0-rc1-work+ #3 Bochs Bochs RIP: 0010:[<ffffffff810b3db0>] [<ffffffff810b3db0>] worker_thread+0x360/0x500 RSP: 0018:ffff88001e1c9de0 EFLAGS: 00010086 RAX: 0000000000000000 RBX: ffff88001e633e00 RCX: 0000000000004140 RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000009 RBP: ffff88001e1c9ea0 R08: 0000000000000000 R09: 0000000000000001 R10: 0000000000000002 R11: 0000000000000000 R12: ffff88001fc8d580 R13: ffff88001fc8d590 R14: ffff88001e633e20 R15: ffff88001e1c6900 FS: 0000000000000000(0000) GS:ffff88001fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 0000000000000000 CR3: 00000000130e8000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process kworker/1:1 (pid: 33, threadinfo ffff88001e1c8000, task ffff88001e1c6900) Stack: ffff880000000000 ffff88001e1c9e40 0000000000000001 ffff88001e1c8010 ffff88001e519c78 ffff88001e1c9e58 ffff88001e1c6900 ffff88001e1c6900 ffff88001e1c6900 ffff88001e1c6900 ffff88001fc8d340 ffff88001fc8d340 Call Trace: [<ffffffff810bc16e>] kthread+0xbe/0xd0 [<ffffffff81bd2664>] kernel_thread_helper+0x4/0x10 Code: b1 00 f6 43 48 02 0f 85 91 01 00 00 48 8b 43 38 48 89 df 48 8b 00 48 89 45 90 e8 ac f0 ff ff 3c 01 0f 85 60 01 00 00 48 8b 53 50 <8b> 02 83 e8 01 85 c0 89 02 0f 84 3b 01 00 00 48 8b 43 38 48 8b RIP [<ffffffff810b3db0>] worker_thread+0x360/0x500 RSP <ffff88001e1c9de0> CR2: 0000000000000000 There was no reason to keep WORKER_REBIND on failure in the first place - WORKER_UNBOUND is guaranteed to be set in such cases preventing incorrectly activating concurrency management. Always clear WORKER_REBIND. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: reimplement idle worker rebinding Currently rebind_workers() uses rebinds idle workers synchronously before proceeding to requesting busy workers to rebind. This is necessary because all workers on @worker_pool->idle_list must be bound before concurrency management local wake-ups from the busy workers take place. Unfortunately, the synchronous idle rebinding is quite complicated. This patch reimplements idle rebinding to simplify the code path. Rather than trying to make all idle workers bound before rebinding busy workers, we simply remove all to-be-bound idle workers from the idle list and let them add themselves back after completing rebinding (successful or not). As only workers which finished rebinding can on on the idle worker list, the idle worker list is guaranteed to have only bound workers unless CPU went down again and local wake-ups are safe. After the change, @worker_pool->nr_idle may deviate than the actual number of idle workers on @worker_pool->idle_list. More specifically, nr_idle may be non-zero while ->idle_list is empty. All users of ->nr_idle and ->idle_list are audited. The only affected one is too_many_workers() which is updated to check %false if ->idle_list is empty regardless of ->nr_idle. After this patch, rebind_workers() no longer performs the nasty idle-rebind retries which require temporary release of gcwq->lock, and both unbinding and rebinding are atomic w.r.t. global_cwq->lock. worker->idle_rebind and global_cwq->rebind_hold are now unnecessary and removed along with the definition of struct idle_rebind. Changed from V1: 1) remove unlikely from too_many_workers(), ->idle_list can be empty anytime, even before this patch, no reason to use unlikely. 2) fix a small rebasing mistake. (which is from rebasing the orignal fixing patch to for-next) 3) add a lot of comments. 4) clear WORKER_REBIND unconditionaly in idle_worker_rebind() tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: WORKER_REBIND is no longer necessary for busy rebinding Because the old unbind/rebinding implementation wasn't atomic w.r.t. GCWQ_DISASSOCIATED manipulation which is protected by global_cwq->lock, we had to use two flags, WORKER_UNBOUND and WORKER_REBIND, to avoid incorrectly losing all NOT_RUNNING bits with back-to-back CPU hotplug operations; otherwise, completion of rebinding while another unbinding is in progress could clear UNBIND prematurely. Now that both unbind/rebinding are atomic w.r.t. GCWQ_DISASSOCIATED, there's no need to use two flags. Just one is enough. Don't use WORKER_REBIND for busy rebinding. tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: WORKER_REBIND is no longer necessary for idle rebinding Now both worker destruction and idle rebinding remove the worker from idle list while it's still idle, so list_empty(&worker->entry) can be used to test whether either is pending and WORKER_DIE to distinguish between the two instead making WORKER_REBIND unnecessary. Use list_empty(&worker->entry) to determine whether destruction or rebinding is pending. This simplifies worker state transitions. WORKER_REBIND is not needed anymore. Remove it. tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: rename manager_mutex to assoc_mutex Now that manager_mutex's role has changed from synchronizing manager role to excluding hotplug against manager, the name is misleading. As it is protecting the CPU-association of the gcwq now, rename it to assoc_mutex. This patch is pure rename and doesn't introduce any functional change. tj: Updated comments and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use __cpuinit instead of __devinit for cpu callbacks For workqueue hotplug callbacks, it makes less sense to use __devinit which discards the memory after boot if !HOTPLUG. __cpuinit, which discards the memory after boot if !HOTPLUG_CPU fits better. tj: Updated description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: fix possible stall on try_to_grab_pending() of a delayed work item Currently, when try_to_grab_pending() grabs a delayed work item, it leaves its linked work items alone on the delayed_works. The linked work items are always NO_COLOR and will cause future cwq_activate_first_delayed() increase cwq->nr_active incorrectly, and may cause the whole cwq to stall. For example, state: cwq->max_active = 1, cwq->nr_active = 1 one work in cwq->pool, many in cwq->delayed_works. step1: try_to_grab_pending() removes a work item from delayed_works but leaves its NO_COLOR linked work items on it. step2: Later on, cwq_activate_first_delayed() activates the linked work item increasing ->nr_active. step3: cwq->nr_active = 1, but all activated work items of the cwq are NO_COLOR. When they finish, cwq->nr_active will not be decreased due to NO_COLOR, and no further work items will be activated from cwq->delayed_works. the cwq stalls. Fix it by ensuring the target work item is activated before stealing PENDING in try_to_grab_pending(). This ensures that all the linked work items are activated without incorrectly bumping cwq->nr_active. tj: Updated comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@kernel.org workqueue: reimplement work_on_cpu() using system_wq The existing work_on_cpu() implementation is hugely inefficient. It creates a new kthread, execute that single function and then let the kthread die on each invocation. Now that system_wq can handle concurrent executions, there's no advantage of doing this. Reimplement work_on_cpu() using system_wq which makes it simpler and way more efficient. stable: While this isn't a fix in itself, it's needed to fix a workqueue related bug in cpufreq/powernow-k8. AFAICS, this shouldn't break other existing users. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Jiri Kosina <jkosina@suse.cz> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Bjorn Helgaas <bhelgaas@google.com> Cc: Len Brown <lenb@kernel.org> Cc: Rafael J. Wysocki <rjw@sisk.pl> Cc: stable@vger.kernel.org workqueue: introduce cwq_set_max_active() helper for thaw_workqueues() Using a helper instead of open code makes thaw_workqueues() clearer. The helper will also be used by the next patch. tj: Slight update to comment and description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: use cwq_set_max_active() helper for workqueue_set_max_active() workqueue_set_max_active() may increase ->max_active without activating delayed works and may make the activation order differ from the queueing order. Both aren't strictly bugs but the resulting behavior could be a bit odd. To make things more consistent, use cwq_set_max_active() helper which immediately makes use of the newly increased max_mactive if there are delayed work items and also keeps the activation order. tj: Slight update to description. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: remove spurious WARN_ON_ONCE(in_irq()) from try_to_grab_pending() e0aecdd874 ("workqueue: use irqsafe timer for delayed_work") made try_to_grab_pending() safe to use from irq context but forgot to remove WARN_ON_ONCE(in_irq()). Remove it. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Fengguang Wu <fengguang.wu@intel.com> workqueue: cancel_delayed_work() should return %false if work item is idle 57b30ae77b ("workqueue: reimplement cancel_delayed_work() using try_to_grab_pending()") made cancel_delayed_work() always return %true unless someone else is also trying to cancel the work item, which is broken - if the target work item is idle, the return value should be %false. try_to_grab_pending() indicates that the target work item was idle by zero return value. Use it for return. Note that this brings cancel_delayed_work() in line with __cancel_work_timer() in return value handling. Signed-off-by: Dan Magenheimer <dan.magenheimer@oracle.com> Signed-off-by: Tejun Heo <tj@kernel.org> LKML-Reference: <444a6439-b1a4-4740-9e7e-bc37267cfe73@default> workqueue: exit rescuer_thread() as TASK_RUNNING A rescue thread exiting TASK_INTERRUPTIBLE can lead to a task scheduling off, never to be seen again. In the case where this occurred, an exiting thread hit reiserfs homebrew conditional resched while holding a mutex, bringing the box to its knees. PID: 18105 TASK: ffff8807fd412180 CPU: 5 COMMAND: "kdmflush" #0 [ffff8808157e7670] schedule at ffffffff8143f489 #1 [ffff8808157e77b8] reiserfs_get_block at ffffffffa038ab2d [reiserfs] #2 [ffff8808157e79a8] __block_write_begin at ffffffff8117fb14 #3 [ffff8808157e7a98] reiserfs_write_begin at ffffffffa0388695 [reiserfs] #4 [ffff8808157e7ad8] generic_perform_write at ffffffff810ee9e2 #5 [ffff8808157e7b58] generic_file_buffered_write at ffffffff810eeb41 #6 [ffff8808157e7ba8] __generic_file_aio_write at ffffffff810f1a3a #7 [ffff8808157e7c58] generic_file_aio_write at ffffffff810f1c88 #8 [ffff8808157e7cc8] do_sync_write at ffffffff8114f850 #9 [ffff8808157e7dd8] do_acct_process at ffffffff810a268f [exception RIP: kernel_thread_helper] RIP: ffffffff8144a5c0 RSP: ffff8808157e7f58 RFLAGS: 00000202 RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffffffff8107af60 RDI: ffff8803ee491d18 RBP: 0000000000000000 R8: 0000000000000000 R9: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 Signed-off-by: Mike Galbraith <mgalbraith@suse.de> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org workqueue: mod_delayed_work_on() shouldn't queue timer on 0 delay 8376fe22c7 ("workqueue: implement mod_delayed_work[_on]()") implemented mod_delayed_work[_on]() using the improved try_to_grab_pending(). The function is later used, among others, to replace [__]candel_delayed_work() + queue_delayed_work() combinations. Unfortunately, a delayed_work item w/ zero @delay is handled slightly differently by mod_delayed_work_on() compared to queue_delayed_work_on(). The latter skips timer altogether and directly queues it using queue_work_on() while the former schedules timer which will expire on the closest tick. This means, when @delay is zero, that [__]cancel_delayed_work() + queue_delayed_work_on() makes the target item immediately executable while mod_delayed_work_on() may induce delay of upto a full tick. This somewhat subtle difference breaks some of the converted users. e.g. block queue plugging uses delayed_work for deferred processing and uses mod_delayed_work_on() when the queue needs to be immediately unplugged. The above problem manifested as noticeably higher number of context switches under certain circumstances. The difference in behavior was caused by missing special case handling for 0 delay in mod_delayed_work_on() compared to queue_delayed_work_on(). Joonsoo Kim posted a patch to add it - ("workqueue: optimize mod_delayed_work_on() when @delay == 0")[1]. The patch was queued for 3.8 but it was described as optimization and I missed that it was a correctness issue. As both queue_delayed_work_on() and mod_delayed_work_on() use __queue_delayed_work() for queueing, it seems that the better approach is to move the 0 delay special handling to the function instead of duplicating it in mod_delayed_work_on(). Fix the problem by moving 0 delay special case handling from queue_delayed_work_on() to __queue_delayed_work(). This replaces Joonsoo's patch. [1] http://thread.gmane.org/gmane.linux.kernel/1379011/focus=1379012 Signed-off-by: Tejun Heo <tj@kernel.org> Reported-and-tested-by: Anders Kaseorg <andersk@MIT.EDU> Reported-and-tested-by: Zlatko Calusic <zlatko.calusic@iskon.hr> LKML-Reference: <alpine.DEB.2.00.1211280953350.26602@dr-wily.mit.edu> LKML-Reference: <50A78AA9.5040904@iskon.hr> Cc: Joonsoo Kim <js1304@gmail.com> workqueue: trivial fix for return statement in work_busy() Return type of work_busy() is unsigned int. There is return statement returning boolean value, 'false' in work_busy(). It is not problem, because 'false' may be treated '0'. However, fixing it would make code robust. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: add WARN_ON_ONCE() on CPU number to wq_worker_waking_up() Recently, workqueue code has gone through some changes and we found some bugs related to concurrency management operations happening on the wrong CPU. When a worker is concurrency managed (!WORKER_NOT_RUNNIG), it should be bound to its associated cpu and woken up to that cpu. Add WARN_ON_ONCE() to verify this. Signed-off-by: Joonsoo Kim <js1304@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> workqueue: convert BUG_ON()s in __queue_delayed_work() to WARN_ON_ONCE()s 8852aac25e ("workqueue: mod_delayed_work_on() shouldn't queue timer on 0 delay") unexpectedly uncovered a very nasty abuse of delayed_work in megaraid - it allocated work_struct, casted it to delayed_work and then pass that into queue_delayed_work(). Previously, this was okay because 0 @delay short-circuited to queue_work() before doing anything with delayed_work. 8852aac25e moved 0 @delay test into __queue_delayed_work() after sanity check on delayed_work making megaraid trigger BUG_ON(). Although megaraid is already fixed by c1d390d8e6 ("megaraid: fix BUG_ON() from incorrect use of delayed work"), this patch converts BUG_ON()s in __queue_delayed_work() to WARN_ON_ONCE()s so that such abusers, if there are more, trigger warning but don't crash the machine. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Xiaotian Feng <xtfeng@gmail.com> wq Change-Id: Ia3c507777a995f32bf6b40dc8318203e53134229 Signed-off-by: franciscofranco <franciscofranco.1990@gmail.com>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Jun 21, 2016
This moves ARM over to the asm-generic/unaligned.h header. This has the benefit of better code generated especially for ARMv7 on gcc 4.7+ compilers. As Arnd Bergmann, points out: The asm-generic version uses the "struct" version for native-endian unaligned access and the "byteshift" version for the opposite endianess. The current ARM version however uses the "byteshift" implementation for both. Thanks to Nicolas Pitre for the excellent analysis: Test case: int foo (int *x) { return get_unaligned(x); } long long bar (long long *x) { return get_unaligned(x); } With the current ARM version: foo: ldrb r3, [r0, stratosk#2] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 2B], MEM[(const u8 *)x_1(D) + 2B] ldrb r1, [r0, stratosk#1] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 1B], MEM[(const u8 *)x_1(D) + 1B] ldrb r2, [r0, #0] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D)], MEM[(const u8 *)x_1(D)] mov r3, r3, asl #16 @ tmp154, MEM[(const u8 *)x_1(D) + 2B], ldrb r0, [r0, stratosk#3] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 3B], MEM[(const u8 *)x_1(D) + 3B] orr r3, r3, r1, asl aosp-mirror#8 @, tmp155, tmp154, MEM[(const u8 *)x_1(D) + 1B], orr r3, r3, r2 @ tmp157, tmp155, MEM[(const u8 *)x_1(D)] orr r0, r3, r0, asl #24 @,, tmp157, MEM[(const u8 *)x_1(D) + 3B], bx lr @ bar: stmfd sp!, {r4, r5, r6, r7} @, mov r2, #0 @ tmp184, ldrb r5, [r0, aosp-mirror#6] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 6B], MEM[(const u8 *)x_1(D) + 6B] ldrb r4, [r0, aosp-mirror#5] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 5B], MEM[(const u8 *)x_1(D) + 5B] ldrb ip, [r0, stratosk#2] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 2B], MEM[(const u8 *)x_1(D) + 2B] ldrb r1, [r0, aosp-mirror#4] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 4B], MEM[(const u8 *)x_1(D) + 4B] mov r5, r5, asl #16 @ tmp175, MEM[(const u8 *)x_1(D) + 6B], ldrb r7, [r0, stratosk#1] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 1B], MEM[(const u8 *)x_1(D) + 1B] orr r5, r5, r4, asl aosp-mirror#8 @, tmp176, tmp175, MEM[(const u8 *)x_1(D) + 5B], ldrb r6, [r0, aosp-mirror#7] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 7B], MEM[(const u8 *)x_1(D) + 7B] orr r5, r5, r1 @ tmp178, tmp176, MEM[(const u8 *)x_1(D) + 4B] ldrb r4, [r0, #0] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D)], MEM[(const u8 *)x_1(D)] mov ip, ip, asl #16 @ tmp188, MEM[(const u8 *)x_1(D) + 2B], ldrb r1, [r0, stratosk#3] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 3B], MEM[(const u8 *)x_1(D) + 3B] orr ip, ip, r7, asl aosp-mirror#8 @, tmp189, tmp188, MEM[(const u8 *)x_1(D) + 1B], orr r3, r5, r6, asl #24 @,, tmp178, MEM[(const u8 *)x_1(D) + 7B], orr ip, ip, r4 @ tmp191, tmp189, MEM[(const u8 *)x_1(D)] orr ip, ip, r1, asl #24 @, tmp194, tmp191, MEM[(const u8 *)x_1(D) + 3B], mov r1, r3 @, orr r0, r2, ip @ tmp171, tmp184, tmp194 ldmfd sp!, {r4, r5, r6, r7} bx lr In both cases the code is slightly suboptimal. One may wonder why wasting r2 with the constant 0 in the second case for example. And all the mov's could be folded in subsequent orr's, etc. Now with the asm-generic version: foo: ldr r0, [r0, #0] @ unaligned @,* x bx lr @ bar: mov r3, r0 @ x, x ldr r0, [r0, #0] @ unaligned @,* x ldr r1, [r3, aosp-mirror#4] @ unaligned @, bx lr @ This is way better of course, but only because this was compiled for ARMv7. In this case the compiler knows that the hardware can do unaligned word access. This isn't that obvious for foo(), but if we remove the get_unaligned() from bar as follows: long long bar (long long *x) {return *x; } then the resulting code is: bar: ldmia r0, {r0, r1} @ x,, bx lr @ So this proves that the presumed aligned vs unaligned cases does have influence on the instructions the compiler may use and that the above unaligned code results are not just an accident. Still... this isn't fully conclusive without at least looking at the resulting assembly fron a pre ARMv6 compilation. Let's see with an ARMv5 target: foo: ldrb r3, [r0, #0] @ zero_extendqisi2 @ tmp139,* x ldrb r1, [r0, stratosk#1] @ zero_extendqisi2 @ tmp140, ldrb r2, [r0, stratosk#2] @ zero_extendqisi2 @ tmp143, ldrb r0, [r0, stratosk#3] @ zero_extendqisi2 @ tmp146, orr r3, r3, r1, asl aosp-mirror#8 @, tmp142, tmp139, tmp140, orr r3, r3, r2, asl #16 @, tmp145, tmp142, tmp143, orr r0, r3, r0, asl #24 @,, tmp145, tmp146, bx lr @ bar: stmfd sp!, {r4, r5, r6, r7} @, ldrb r2, [r0, #0] @ zero_extendqisi2 @ tmp139,* x ldrb r7, [r0, stratosk#1] @ zero_extendqisi2 @ tmp140, ldrb r3, [r0, aosp-mirror#4] @ zero_extendqisi2 @ tmp149, ldrb r6, [r0, aosp-mirror#5] @ zero_extendqisi2 @ tmp150, ldrb r5, [r0, stratosk#2] @ zero_extendqisi2 @ tmp143, ldrb r4, [r0, aosp-mirror#6] @ zero_extendqisi2 @ tmp153, ldrb r1, [r0, aosp-mirror#7] @ zero_extendqisi2 @ tmp156, ldrb ip, [r0, stratosk#3] @ zero_extendqisi2 @ tmp146, orr r2, r2, r7, asl aosp-mirror#8 @, tmp142, tmp139, tmp140, orr r3, r3, r6, asl aosp-mirror#8 @, tmp152, tmp149, tmp150, orr r2, r2, r5, asl #16 @, tmp145, tmp142, tmp143, orr r3, r3, r4, asl #16 @, tmp155, tmp152, tmp153, orr r0, r2, ip, asl #24 @,, tmp145, tmp146, orr r1, r3, r1, asl #24 @,, tmp155, tmp156, ldmfd sp!, {r4, r5, r6, r7} bx lr Compared to the initial results, this is really nicely optimized and I couldn't do much better if I were to hand code it myself. Signed-off-by: Rob Herring <rob.herring@calxeda.com> Reviewed-by: Nicolas Pitre <nico@linaro.org> Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> modified for Mako from kernel.org reference Signed-off-by: faux123 <reioux@gmail.com>
stratosk
pushed a commit
that referenced
this pull request
Aug 11, 2016
commit 75a6f82 upstream. Normally opening a file, unlinking it and then closing will have the inode freed upon close() (provided that it's not otherwise busy and has no remaining links, of course). However, there's one case where that does *not* happen. Namely, if you open it by fhandle with cold dcache, then unlink() and close(). In normal case you get d_delete() in unlink(2) notice that dentry is busy and unhash it; on the final dput() it will be forcibly evicted from dcache, triggering iput() and inode removal. In this case, though, we end up with *two* dentries - disconnected (created by open-by-fhandle) and regular one (used by unlink()). The latter will have its reference to inode dropped just fine, but the former will not - it's considered hashed (it is on the ->s_anon list), so it will stay around until the memory pressure will finally do it in. As the result, we have the final iput() delayed indefinitely. It's trivial to reproduce - void flush_dcache(void) { system("mount -o remount,rw /"); } static char buf[20 * 1024 * 1024]; main() { int fd; union { struct file_handle f; char buf[MAX_HANDLE_SZ]; } x; int m; x.f.handle_bytes = sizeof(x); chdir("/root"); mkdir("foo", 0700); fd = open("foo/bar", O_CREAT | O_RDWR, 0600); close(fd); name_to_handle_at(AT_FDCWD, "foo/bar", &x.f, &m, 0); flush_dcache(); fd = open_by_handle_at(AT_FDCWD, &x.f, O_RDWR); unlink("foo/bar"); write(fd, buf, sizeof(buf)); system("df ."); /* 20Mb eaten */ close(fd); system("df ."); /* should've freed those 20Mb */ flush_dcache(); system("df ."); /* should be the same as #2 */ } will spit out something like Filesystem 1K-blocks Used Available Use% Mounted on /dev/root 322023 303843 1131 100% / Filesystem 1K-blocks Used Available Use% Mounted on /dev/root 322023 303843 1131 100% / Filesystem 1K-blocks Used Available Use% Mounted on /dev/root 322023 283282 21692 93% / - inode gets freed only when dentry is finally evicted (here we trigger than by remount; normally it would've happened in response to memory pressure hell knows when). Acked-by: J. Bruce Fields <bfields@fieldses.org> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Aug 11, 2016
commit ecf5fc6 upstream. Nikolay has reported a hang when a memcg reclaim got stuck with the following backtrace: PID: 18308 TASK: ffff883d7c9b0a30 CPU: 1 COMMAND: "rsync" #0 __schedule at ffffffff815ab152 #1 schedule at ffffffff815ab76e #2 schedule_timeout at ffffffff815ae5e5 #3 io_schedule_timeout at ffffffff815aad6a aosp-mirror#4 bit_wait_io at ffffffff815abfc6 aosp-mirror#5 __wait_on_bit at ffffffff815abda5 aosp-mirror#6 wait_on_page_bit at ffffffff8111fd4f aosp-mirror#7 shrink_page_list at ffffffff81135445 aosp-mirror#8 shrink_inactive_list at ffffffff81135845 aosp-mirror#9 shrink_lruvec at ffffffff81135ead #10 shrink_zone at ffffffff811360c3 #11 shrink_zones at ffffffff81136eff #12 do_try_to_free_pages at ffffffff8113712f #13 try_to_free_mem_cgroup_pages at ffffffff811372be #14 try_charge at ffffffff81189423 #15 mem_cgroup_try_charge at ffffffff8118c6f5 #16 __add_to_page_cache_locked at ffffffff8112137d #17 add_to_page_cache_lru at ffffffff81121618 #18 pagecache_get_page at ffffffff8112170b #19 grow_dev_page at ffffffff811c8297 #20 __getblk_slow at ffffffff811c91d6 #21 __getblk_gfp at ffffffff811c92c1 #22 ext4_ext_grow_indepth at ffffffff8124565c #23 ext4_ext_create_new_leaf at ffffffff81246ca8 #24 ext4_ext_insert_extent at ffffffff81246f09 #25 ext4_ext_map_blocks at ffffffff8124a848 #26 ext4_map_blocks at ffffffff8121a5b7 #27 mpage_map_one_extent at ffffffff8121b1fa #28 mpage_map_and_submit_extent at ffffffff8121f07b #29 ext4_writepages at ffffffff8121f6d5 #30 do_writepages at ffffffff8112c490 #31 __filemap_fdatawrite_range at ffffffff81120199 #32 filemap_flush at ffffffff8112041c #33 ext4_alloc_da_blocks at ffffffff81219da1 #34 ext4_rename at ffffffff81229b91 #35 ext4_rename2 at ffffffff81229e32 #36 vfs_rename at ffffffff811a08a5 #37 SYSC_renameat2 at ffffffff811a3ffc #38 sys_renameat2 at ffffffff811a408e #39 sys_rename at ffffffff8119e51e #40 system_call_fastpath at ffffffff815afa89 Dave Chinner has properly pointed out that this is a deadlock in the reclaim code because ext4 doesn't submit pages which are marked by PG_writeback right away. The heuristic was introduced by commit e62e384 ("memcg: prevent OOM with too many dirty pages") and it was applied only when may_enter_fs was specified. The code has been changed by c3b94f4 ("memcg: further prevent OOM with too many dirty pages") which has removed the __GFP_FS restriction with a reasoning that we do not get into the fs code. But this is not sufficient apparently because the fs doesn't necessarily submit pages marked PG_writeback for IO right away. ext4_bio_write_page calls io_submit_add_bh but that doesn't necessarily submit the bio. Instead it tries to map more pages into the bio and mpage_map_one_extent might trigger memcg charge which might end up waiting on a page which is marked PG_writeback but hasn't been submitted yet so we would end up waiting for something that never finishes. Fix this issue by replacing __GFP_IO by may_enter_fs check (for case 2) before we go to wait on the writeback. The page fault path, which is the only path that triggers memcg oom killer since 3.12, shouldn't require GFP_NOFS and so we shouldn't reintroduce the premature OOM killer issue which was originally addressed by the heuristic. As per David Chinner the xfs is doing similar thing since 2.6.15 already so ext4 is not the only affected filesystem. Moreover he notes: : For example: IO completion might require unwritten extent conversion : which executes filesystem transactions and GFP_NOFS allocations. The : writeback flag on the pages can not be cleared until unwritten : extent conversion completes. Hence memory reclaim cannot wait on : page writeback to complete in GFP_NOFS context because it is not : safe to do so, memcg reclaim or otherwise. [tytso@mit.edu: corrected the control flow] Fixes: c3b94f4 ("memcg: further prevent OOM with too many dirty pages") Reported-by: Nikolay Borisov <kernel@kyup.com> Signed-off-by: Michal Hocko <mhocko@suse.cz> Signed-off-by: Hugh Dickins <hughd@google.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Aug 11, 2016
[ Upstream commit 0470eb9 ] Kirill A. Shutemov says: This simple test-case trigers few locking asserts in kernel: int main(int argc, char **argv) { unsigned int block_size = 16 * 4096; struct nl_mmap_req req = { .nm_block_size = block_size, .nm_block_nr = 64, .nm_frame_size = 16384, .nm_frame_nr = 64 * block_size / 16384, }; unsigned int ring_size; int fd; fd = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC); if (setsockopt(fd, SOL_NETLINK, NETLINK_RX_RING, &req, sizeof(req)) < 0) exit(1); if (setsockopt(fd, SOL_NETLINK, NETLINK_TX_RING, &req, sizeof(req)) < 0) exit(1); ring_size = req.nm_block_nr * req.nm_block_size; mmap(NULL, 2 * ring_size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0); return 0; } +++ exited with 0 +++ BUG: sleeping function called from invalid context at /home/kas/git/public/linux-mm/kernel/locking/mutex.c:616 in_atomic(): 1, irqs_disabled(): 0, pid: 1, name: init 3 locks held by init/1: #0: (reboot_mutex){+.+...}, at: [<ffffffff81080959>] SyS_reboot+0xa9/0x220 #1: ((reboot_notifier_list).rwsem){.+.+..}, at: [<ffffffff8107f379>] __blocking_notifier_call_chain+0x39/0x70 #2: (rcu_callback){......}, at: [<ffffffff810d32e0>] rcu_do_batch.isra.49+0x160/0x10c0 Preemption disabled at:[<ffffffff8145365f>] __delay+0xf/0x20 CPU: 1 PID: 1 Comm: init Not tainted 4.1.0-00009-gbddf4c4818e0 #253 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS Debian-1.8.2-1 04/01/2014 ffff88017b3d8000 ffff88027bc03c38 ffffffff81929ceb 0000000000000102 0000000000000000 ffff88027bc03c68 ffffffff81085a9d 0000000000000002 ffffffff81ca2a20 0000000000000268 0000000000000000 ffff88027bc03c98 Call Trace: <IRQ> [<ffffffff81929ceb>] dump_stack+0x4f/0x7b [<ffffffff81085a9d>] ___might_sleep+0x16d/0x270 [<ffffffff81085bed>] __might_sleep+0x4d/0x90 [<ffffffff8192e96f>] mutex_lock_nested+0x2f/0x430 [<ffffffff81932fed>] ? _raw_spin_unlock_irqrestore+0x5d/0x80 [<ffffffff81464143>] ? __this_cpu_preempt_check+0x13/0x20 [<ffffffff8182fc3d>] netlink_set_ring+0x1ed/0x350 [<ffffffff8182e000>] ? netlink_undo_bind+0x70/0x70 [<ffffffff8182fe20>] netlink_sock_destruct+0x80/0x150 [<ffffffff817e484d>] __sk_free+0x1d/0x160 [<ffffffff817e49a9>] sk_free+0x19/0x20 [..] Cong Wang says: We can't hold mutex lock in a rcu callback, [..] Thomas Graf says: The socket should be dead at this point. It might be simpler to add a netlink_release_ring() function which doesn't require locking at all. Reported-by: "Kirill A. Shutemov" <kirill@shutemov.name> Diagnosed-by: Cong Wang <cwang@twopensource.com> Suggested-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Aug 11, 2016
…_BH() in preemptible context. [ Upstream commit 44f49dd ] Fixes the following kernel BUG : BUG: using __this_cpu_add() in preemptible [00000000] code: bash/2758 caller is __this_cpu_preempt_check+0x13/0x15 CPU: 0 PID: 2758 Comm: bash Tainted: P O 3.18.19 #2 ffffffff8170eaca ffff880110d1b788 ffffffff81482b2a 0000000000000000 0000000000000000 ffff880110d1b7b8 ffffffff812010ae ffff880007cab800 ffff88001a060800 ffff88013a899108 ffff880108b84240 ffff880110d1b7c8 Call Trace: [<ffffffff81482b2a>] dump_stack+0x52/0x80 [<ffffffff812010ae>] check_preemption_disabled+0xce/0xe1 [<ffffffff812010d4>] __this_cpu_preempt_check+0x13/0x15 [<ffffffff81419d60>] ipmr_queue_xmit+0x647/0x70c [<ffffffff8141a154>] ip_mr_forward+0x32f/0x34e [<ffffffff8141af76>] ip_mroute_setsockopt+0xe03/0x108c [<ffffffff810553fc>] ? get_parent_ip+0x11/0x42 [<ffffffff810e6974>] ? pollwake+0x4d/0x51 [<ffffffff81058ac0>] ? default_wake_function+0x0/0xf [<ffffffff810553fc>] ? get_parent_ip+0x11/0x42 [<ffffffff810613d9>] ? __wake_up_common+0x45/0x77 [<ffffffff81486ea9>] ? _raw_spin_unlock_irqrestore+0x1d/0x32 [<ffffffff810618bc>] ? __wake_up_sync_key+0x4a/0x53 [<ffffffff8139a519>] ? sock_def_readable+0x71/0x75 [<ffffffff813dd226>] do_ip_setsockopt+0x9d/0xb55 [<ffffffff81429818>] ? unix_seqpacket_sendmsg+0x3f/0x41 [<ffffffff813963fe>] ? sock_sendmsg+0x6d/0x86 [<ffffffff813959d4>] ? sockfd_lookup_light+0x12/0x5d [<ffffffff8139650a>] ? SyS_sendto+0xf3/0x11b [<ffffffff810d5738>] ? new_sync_read+0x82/0xaa [<ffffffff813ddd19>] compat_ip_setsockopt+0x3b/0x99 [<ffffffff813fb24a>] compat_raw_setsockopt+0x11/0x32 [<ffffffff81399052>] compat_sock_common_setsockopt+0x18/0x1f [<ffffffff813c4d05>] compat_SyS_setsockopt+0x1a9/0x1cf [<ffffffff813c4149>] compat_SyS_socketcall+0x180/0x1e3 [<ffffffff81488ea1>] cstar_dispatch+0x7/0x1e Signed-off-by: Ani Sinha <ani@arista.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Aug 11, 2016
commit 1f9c6e1 upstream. There were several bugs here. 1) The done label was in the wrong place so we didn't copy any information out when there was no command given. 2) We were using PAGE_SIZE as the size of the buffer instead of "PAGE_SIZE - pos". 3) snprintf() returns the number of characters that would have been printed if there were enough space. If there was not enough space (and we had fixed the memory corruption bug #2) then it would result in an information leak when we do simple_read_from_buffer(). I've changed it to use scnprintf() instead. I also removed the initialization at the start of the function, because I thought it made the code a little more clear. Fixes: 5e6e3a9 ('wireless: mwifiex: initial commit for Marvell mwifiex driver') Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Acked-by: Amitkumar Karwar <akarwar@marvell.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Aug 11, 2016
[ Upstream commit 1b8e6a0 ] When a passive TCP is created, we eventually call tcp_md5_do_add() with sk pointing to the child. It is not owner by the user yet (we will add this socket into listener accept queue a bit later anyway) But we do own the spinlock, so amend the lockdep annotation to avoid following splat : [ 8451.090932] net/ipv4/tcp_ipv4.c:923 suspicious rcu_dereference_protected() usage! [ 8451.090932] [ 8451.090932] other info that might help us debug this: [ 8451.090932] [ 8451.090934] [ 8451.090934] rcu_scheduler_active = 1, debug_locks = 1 [ 8451.090936] 3 locks held by socket_sockopt_/214795: [ 8451.090936] #0: (rcu_read_lock){.+.+..}, at: [<ffffffff855c6ac1>] __netif_receive_skb_core+0x151/0xe90 [ 8451.090947] #1: (rcu_read_lock){.+.+..}, at: [<ffffffff85618143>] ip_local_deliver_finish+0x43/0x2b0 [ 8451.090952] #2: (slock-AF_INET){+.-...}, at: [<ffffffff855acda5>] sk_clone_lock+0x1c5/0x500 [ 8451.090958] [ 8451.090958] stack backtrace: [ 8451.090960] CPU: 7 PID: 214795 Comm: socket_sockopt_ [ 8451.091215] Call Trace: [ 8451.091216] <IRQ> [<ffffffff856fb29c>] dump_stack+0x55/0x76 [ 8451.091229] [<ffffffff85123b5b>] lockdep_rcu_suspicious+0xeb/0x110 [ 8451.091235] [<ffffffff8564544f>] tcp_md5_do_add+0x1bf/0x1e0 [ 8451.091239] [<ffffffff85645751>] tcp_v4_syn_recv_sock+0x1f1/0x4c0 [ 8451.091242] [<ffffffff85642b27>] ? tcp_v4_md5_hash_skb+0x167/0x190 [ 8451.091246] [<ffffffff85647c78>] tcp_check_req+0x3c8/0x500 [ 8451.091249] [<ffffffff856451ae>] ? tcp_v4_inbound_md5_hash+0x11e/0x190 [ 8451.091253] [<ffffffff85647170>] tcp_v4_rcv+0x3c0/0x9f0 [ 8451.091256] [<ffffffff85618143>] ? ip_local_deliver_finish+0x43/0x2b0 [ 8451.091260] [<ffffffff856181b6>] ip_local_deliver_finish+0xb6/0x2b0 [ 8451.091263] [<ffffffff85618143>] ? ip_local_deliver_finish+0x43/0x2b0 [ 8451.091267] [<ffffffff85618d38>] ip_local_deliver+0x48/0x80 [ 8451.091270] [<ffffffff85618510>] ip_rcv_finish+0x160/0x700 [ 8451.091273] [<ffffffff8561900e>] ip_rcv+0x29e/0x3d0 [ 8451.091277] [<ffffffff855c74b7>] __netif_receive_skb_core+0xb47/0xe90 Fixes: a8afca0 ("tcp: md5: protects md5sig_info with RCU") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Aug 11, 2016
commit e46e31a upstream. When using the Promise TX2+ SATA controller on PA-RISC, the system often crashes with kernel panic, for example just writing data with the dd utility will make it crash. Kernel panic - not syncing: drivers/parisc/sba_iommu.c: I/O MMU @ 000000000000a000 is out of mapping resources CPU: 0 PID: 18442 Comm: mkspadfs Not tainted 4.4.0-rc2 #2 Backtrace: [<000000004021497c>] show_stack+0x14/0x20 [<0000000040410bf0>] dump_stack+0x88/0x100 [<000000004023978c>] panic+0x124/0x360 [<0000000040452c18>] sba_alloc_range+0x698/0x6a0 [<0000000040453150>] sba_map_sg+0x260/0x5b8 [<000000000c18dbb4>] ata_qc_issue+0x264/0x4a8 [libata] [<000000000c19535c>] ata_scsi_translate+0xe4/0x220 [libata] [<000000000c19a93c>] ata_scsi_queuecmd+0xbc/0x320 [libata] [<0000000040499bbc>] scsi_dispatch_cmd+0xfc/0x130 [<000000004049da34>] scsi_request_fn+0x6e4/0x970 [<00000000403e95a8>] __blk_run_queue+0x40/0x60 [<00000000403e9d8c>] blk_run_queue+0x3c/0x68 [<000000004049a534>] scsi_run_queue+0x2a4/0x360 [<000000004049be68>] scsi_end_request+0x1a8/0x238 [<000000004049de84>] scsi_io_completion+0xfc/0x688 [<0000000040493c74>] scsi_finish_command+0x17c/0x1d0 The cause of the crash is not exhaustion of the IOMMU space, there is plenty of free pages. The function sba_alloc_range is called with size 0x11000, thus the pages_needed variable is 0x11. The function sba_search_bitmap is called with bits_wanted 0x11 and boundary size is 0x10 (because dma_get_seg_boundary(dev) returns 0xffff). The function sba_search_bitmap attempts to allocate 17 pages that must not cross 16-page boundary - it can't satisfy this requirement (iommu_is_span_boundary always returns true) and fails even if there are many free entries in the IOMMU space. How did it happen that we try to allocate 17 pages that don't cross 16-page boundary? The cause is in the function iommu_coalesce_chunks. This function tries to coalesce adjacent entries in the scatterlist. The function does several checks if it may coalesce one entry with the next, one of those checks is this: if (startsg->length + dma_len > max_seg_size) break; When it finishes coalescing adjacent entries, it allocates the mapping: sg_dma_len(contig_sg) = dma_len; dma_len = ALIGN(dma_len + dma_offset, IOVP_SIZE); sg_dma_address(contig_sg) = PIDE_FLAG | (iommu_alloc_range(ioc, dev, dma_len) << IOVP_SHIFT) | dma_offset; It is possible that (startsg->length + dma_len > max_seg_size) is false (we are just near the 0x10000 max_seg_size boundary), so the funcion decides to coalesce this entry with the next entry. When the coalescing succeeds, the function performs dma_len = ALIGN(dma_len + dma_offset, IOVP_SIZE); And now, because of non-zero dma_offset, dma_len is greater than 0x10000. iommu_alloc_range (a pointer to sba_alloc_range) is called and it attempts to allocate 17 pages for a device that must not cross 16-page boundary. To fix the bug, we must make sure that dma_len after addition of dma_offset and alignment doesn't cross the segment boundary. I.e. change if (startsg->length + dma_len > max_seg_size) break; to if (ALIGN(dma_len + dma_offset + startsg->length, IOVP_SIZE) > max_seg_size) break; This patch makes this change (it precalculates max_seg_boundary at the beginning of the function iommu_coalesce_chunks). I also added a check that the mapping length doesn't exceed dma_get_seg_boundary(dev) (it is not needed for Promise TX2+ SATA, but it may be needed for other devices that have dma_get_seg_boundary lower than dma_get_max_seg_size). Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Helge Deller <deller@gmx.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Aug 11, 2016
commit 00cd29b upstream. The starting node for a klist iteration is often passed in from somewhere way above the klist infrastructure, meaning there's no guarantee the node is still on the list. We've seen this in SCSI where we use bus_find_device() to iterate through a list of devices. In the face of heavy hotplug activity, the last device returned by bus_find_device() can be removed before the next call. This leads to Dec 3 13:22:02 localhost kernel: WARNING: CPU: 2 PID: 28073 at include/linux/kref.h:47 klist_iter_init_node+0x3d/0x50() Dec 3 13:22:02 localhost kernel: Modules linked in: scsi_debug x86_pkg_temp_thermal kvm_intel kvm irqbypass crc32c_intel joydev iTCO_wdt dcdbas ipmi_devintf acpi_power_meter iTCO_vendor_support ipmi_si imsghandler pcspkr wmi acpi_cpufreq tpm_tis tpm shpchp lpc_ich mfd_core nfsd nfs_acl lockd grace sunrpc tg3 ptp pps_core Dec 3 13:22:02 localhost kernel: CPU: 2 PID: 28073 Comm: cat Not tainted 4.4.0-rc1+ #2 Dec 3 13:22:02 localhost kernel: Hardware name: Dell Inc. PowerEdge R320/08VT7V, BIOS 2.0.22 11/19/2013 Dec 3 13:22:02 localhost kernel: ffffffff81a20e77 ffff880613acfd18 ffffffff81321eef 0000000000000000 Dec 3 13:22:02 localhost kernel: ffff880613acfd50 ffffffff8107ca52 ffff88061176b198 0000000000000000 Dec 3 13:22:02 localhost kernel: ffffffff814542b0 ffff880610cfb100 ffff88061176b198 ffff880613acfd60 Dec 3 13:22:02 localhost kernel: Call Trace: Dec 3 13:22:02 localhost kernel: [<ffffffff81321eef>] dump_stack+0x44/0x55 Dec 3 13:22:02 localhost kernel: [<ffffffff8107ca52>] warn_slowpath_common+0x82/0xc0 Dec 3 13:22:02 localhost kernel: [<ffffffff814542b0>] ? proc_scsi_show+0x20/0x20 Dec 3 13:22:02 localhost kernel: [<ffffffff8107cb4a>] warn_slowpath_null+0x1a/0x20 Dec 3 13:22:02 localhost kernel: [<ffffffff8167225d>] klist_iter_init_node+0x3d/0x50 Dec 3 13:22:02 localhost kernel: [<ffffffff81421d41>] bus_find_device+0x51/0xb0 Dec 3 13:22:02 localhost kernel: [<ffffffff814545ad>] scsi_seq_next+0x2d/0x40 [...] And an eventual crash. It can actually occur in any hotplug system which has a device finder and a starting device. We can fix this globally by making sure the starting node for klist_iter_init_node() is actually a member of the list before using it (and by starting from the beginning if it isn't). Reported-by: Ewan D. Milne <emilne@redhat.com> Tested-by: Ewan D. Milne <emilne@redhat.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Aug 11, 2016
commit 361cad3 upstream. We've seen this in a packet capture - I've intermixed what I think was going on. The fix here is to grab the so_lock sooner. 1964379 -> #1 open (for write) reply seqid=1 1964393 -> #2 open (for read) reply seqid=2 __nfs4_close(), state->n_wronly-- nfs4_state_set_mode_locked(), changes state->state = [R] state->flags is [RW] state->state is [R], state->n_wronly == 0, state->n_rdonly == 1 1964398 -> #3 open (for write) call -> because close is already running 1964399 -> downgrade (to read) call seqid=2 (close of #1) 1964402 -> #3 open (for write) reply seqid=3 __update_open_stateid() nfs_set_open_stateid_locked(), changes state->flags state->flags is [RW] state->state is [R], state->n_wronly == 0, state->n_rdonly == 1 new sequence number is exposed now via nfs4_stateid_copy() next step would be update_open_stateflags(), pending so_lock 1964403 -> downgrade reply seqid=2, fails with OLD_STATEID (close of #1) nfs4_close_prepare() gets so_lock and recalcs flags -> send close 1964405 -> downgrade (to read) call seqid=3 (close of #1 retry) __update_open_stateid() gets so_lock * update_open_stateflags() updates state->n_wronly. nfs4_state_set_mode_locked() updates state->state state->flags is [RW] state->state is [RW], state->n_wronly == 1, state->n_rdonly == 1 * should have suppressed the preceding nfs4_close_prepare() from sending open_downgrade 1964406 -> write call 1964408 -> downgrade (to read) reply seqid=4 (close of #1 retry) nfs_clear_open_stateid_locked() state->flags is [R] state->state is [RW], state->n_wronly == 1, state->n_rdonly == 1 1964409 -> write reply (fails, openmode) Signed-off-by: Andrew Elble <aweits@rit.edu> Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stratosk
pushed a commit
that referenced
this pull request
Aug 11, 2016
commit b49b927 upstream. We shouldn't be calling clk_prepare_enable()/clk_prepare_disable() in an atomic context. Fixes the following issue: [ 5.830970] ehci-omap: OMAP-EHCI Host Controller driver [ 5.830974] driver_register 'ehci-omap' [ 5.895849] driver_register 'wl1271_sdio' [ 5.896870] BUG: scheduling while atomic: udevd/994/0x00000002 [ 5.896876] 4 locks held by udevd/994: [ 5.896904] #0: (&dev->mutex){......}, at: [<c049597c>] __driver_attach+0x60/0xac [ 5.896923] #1: (&dev->mutex){......}, at: [<c049598c>] __driver_attach+0x70/0xac [ 5.896946] #2: (tll_lock){+.+...}, at: [<c04c2630>] omap_tll_enable+0x2c/0xd0 [ 5.896966] #3: (prepare_lock){+.+...}, at: [<c05ce9c8>] clk_prepare_lock+0x48/0xe0 [ 5.897042] Modules linked in: wlcore_sdio(+) ehci_omap(+) dwc3_omap snd_soc_ts3a225e leds_is31fl319x bq27xxx_battery_i2c tsc2007 bq27xxx_battery bq2429x_charger ina2xx tca8418_keypad as5013 leds_tca6507 twl6040_vibra gpio_twl6040 bmp085_i2c(+) palmas_gpadc usb3503 palmas_pwrbutton bmg160_i2c(+) bmp085 bma150(+) bmg160_core bmp280 input_polldev snd_soc_omap_mcbsp snd_soc_omap_mcpdm snd_soc_omap snd_pcm_dmaengine [ 5.897048] Preemption disabled at:[< (null)>] (null) [ 5.897051] [ 5.897059] CPU: 0 PID: 994 Comm: udevd Not tainted 4.6.0-rc5-letux+ #233 [ 5.897062] Hardware name: Generic OMAP5 (Flattened Device Tree) [ 5.897076] [<c010e714>] (unwind_backtrace) from [<c010af34>] (show_stack+0x10/0x14) [ 5.897087] [<c010af34>] (show_stack) from [<c040aa7c>] (dump_stack+0x88/0xc0) [ 5.897099] [<c040aa7c>] (dump_stack) from [<c020c558>] (__schedule_bug+0xac/0xd0) [ 5.897111] [<c020c558>] (__schedule_bug) from [<c06f3d44>] (__schedule+0x88/0x7e4) [ 5.897120] [<c06f3d44>] (__schedule) from [<c06f46d8>] (schedule+0x9c/0xc0) [ 5.897129] [<c06f46d8>] (schedule) from [<c06f4904>] (schedule_preempt_disabled+0x14/0x20) [ 5.897140] [<c06f4904>] (schedule_preempt_disabled) from [<c06f64e4>] (mutex_lock_nested+0x258/0x43c) [ 5.897150] [<c06f64e4>] (mutex_lock_nested) from [<c05ce9c8>] (clk_prepare_lock+0x48/0xe0) [ 5.897160] [<c05ce9c8>] (clk_prepare_lock) from [<c05d0e7c>] (clk_prepare+0x10/0x28) [ 5.897169] [<c05d0e7c>] (clk_prepare) from [<c04c2668>] (omap_tll_enable+0x64/0xd0) [ 5.897180] [<c04c2668>] (omap_tll_enable) from [<c04c1728>] (usbhs_runtime_resume+0x18/0x17c) [ 5.897192] [<c04c1728>] (usbhs_runtime_resume) from [<c049d404>] (pm_generic_runtime_resume+0x2c/0x40) [ 5.897202] [<c049d404>] (pm_generic_runtime_resume) from [<c049f180>] (__rpm_callback+0x38/0x68) [ 5.897210] [<c049f180>] (__rpm_callback) from [<c049f220>] (rpm_callback+0x70/0x88) [ 5.897218] [<c049f220>] (rpm_callback) from [<c04a0a00>] (rpm_resume+0x4ec/0x7ec) [ 5.897227] [<c04a0a00>] (rpm_resume) from [<c04a0f48>] (__pm_runtime_resume+0x4c/0x64) [ 5.897236] [<c04a0f48>] (__pm_runtime_resume) from [<c04958dc>] (driver_probe_device+0x30/0x70) [ 5.897246] [<c04958dc>] (driver_probe_device) from [<c04959a4>] (__driver_attach+0x88/0xac) [ 5.897256] [<c04959a4>] (__driver_attach) from [<c04940f8>] (bus_for_each_dev+0x50/0x84) [ 5.897267] [<c04940f8>] (bus_for_each_dev) from [<c0494e40>] (bus_add_driver+0xcc/0x1e4) [ 5.897276] [<c0494e40>] (bus_add_driver) from [<c0496914>] (driver_register+0xac/0xf4) [ 5.897286] [<c0496914>] (driver_register) from [<c01018e0>] (do_one_initcall+0x100/0x1b8) [ 5.897296] [<c01018e0>] (do_one_initcall) from [<c01c7a54>] (do_init_module+0x58/0x1c0) [ 5.897304] [<c01c7a54>] (do_init_module) from [<c01c8a3c>] (SyS_finit_module+0x88/0x90) [ 5.897313] [<c01c8a3c>] (SyS_finit_module) from [<c0107120>] (ret_fast_syscall+0x0/0x1c) [ 5.912697] ------------[ cut here ]------------ [ 5.912711] WARNING: CPU: 0 PID: 994 at kernel/sched/core.c:2996 _raw_spin_unlock+0x28/0x58 [ 5.912717] DEBUG_LOCKS_WARN_ON(val > preempt_count()) Reported-by: H. Nikolaus Schaller <hns@goldelico.com> Tested-by: H. Nikolaus Schaller <hns@goldelico.com> Signed-off-by: Roger Quadros <rogerq@ti.com> Signed-off-by: Lee Jones <lee.jones@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Aug 13, 2016
msm_sat_enqueue() calls spin_lock() and msm_sat_dequeue() calls spin_lock_irqsave(). This leads to lockdep warnings about the same lock being taken in interrupts on and interrupts off context which can lead to a potential deadlock. ================================= [ INFO: inconsistent lock state ] 3.4.0+ #382 Tainted: G W --------------------------------- inconsistent {HARDIRQ-ON-W} -> {IN-HARDIRQ-W} usage. kworker/u:2/94 [HC1[1]:SC0[0]:HE0:SE1] takes: (&(&sat->lock)->rlock){?.+...}, at: [<c0348b80>] msm_sat_enqueue+0x20/0xb0 {HARDIRQ-ON-W} state was registered at: [<c00bd290>] __lock_acquire+0x664/0x8d8 [<c00bd690>] lock_acquire+0x18c/0x1e8 [<c06989a4>] _raw_spin_lock+0x38/0x48 [<c0348b80>] msm_sat_enqueue+0x20/0xb0 [<c0349338>] msm_slim_rxwq+0x278/0x42c [<c0349580>] msm_slim_rx_msgq_thread+0x94/0x1f8 [<c008e480>] kthread+0x90/0xa0 [<c000f438>] kernel_thread_exit+0x0/0x8 irq event stamp: 24219 hardirqs last enabled at (24218): [<c0699188>] _raw_spin_unlock_irqrestore+0x3c/0x68 hardirqs last disabled at (24219): [<c06993f4>] __irq_svc+0x34/0x78 softirqs last enabled at (24129): [<c00738a8>] irq_exit+0x54/0xa8 softirqs last disabled at (24108): [<c00738a8>] irq_exit+0x54/0xa8 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&(&sat->lock)->rlock); <Interrupt> lock(&(&sat->lock)->rlock); *** DEADLOCK *** 3 locks held by kworker/u:2/94: #0: ((sat->satcl.name)){.+.+.+}, at: [<c0087124>] process_one_work+0x1e8/0x648 stratosk#1: ((&sat->wd)){+.+.+.}, at: [<c0087124>] process_one_work+0x1e8/0x648 stratosk#2: (&dev->tx_lock){+.+...}, at: [<c0349cac>] msm_xfer_msg+0xb4/0x51c stack backtrace: [<c00151b0>] (unwind_backtrace+0x0/0x120) from [<c00b9d48>] (print_usage_bug+0x258/0x2c0) [<c00b9d48>] (print_usage_bug+0x258/0x2c0) from [<c00ba12c>] (mark_lock+0x37c/0x68c) [<c00ba12c>] (mark_lock+0x37c/0x68c) from [<c00bd20c>] (__lock_acquire+0x5e0/0x8d8) [<c00bd20c>] (__lock_acquire+0x5e0/0x8d8) from [<c00bd690>] (lock_acquire+0x18c/0x1e8) [<c00bd690>] (lock_acquire+0x18c/0x1e8) from [<c06989a4>] (_raw_spin_lock+0x38/0x48) [<c06989a4>] (_raw_spin_lock+0x38/0x48) from [<c0348b80>] (msm_sat_enqueue+0x20/0xb0) [<c0348b80>] (msm_sat_enqueue+0x20/0xb0) from [<c0348dc8>] (msm_slim_interrupt+0x1b8/0x4b0) [<c0348dc8>] (msm_slim_interrupt+0x1b8/0x4b0) from [<c00d2ccc>] (handle_irq_event_percpu+0x118/0x3a4) [<c00d2ccc>] (handle_irq_event_percpu+0x118/0x3a4) from [<c00d2f94>] (handle_irq_event+0x3c/0x5c) [<c00d2f94>] (handle_irq_event+0x3c/0x5c) from [<c00d5c4c>] (handle_fasteoi_irq+0xd0/0x11c) [<c00d5c4c>] (handle_fasteoi_irq+0xd0/0x11c) from [<c00d2ae4>] (generic_handle_irq+0x24/0x2c) [<c00d2ae4>] (generic_handle_irq+0x24/0x2c) from [<c000f370>] (handle_IRQ+0x7c/0xc0) [<c000f370>] (handle_IRQ+0x7c/0xc0) from [<c0008774>] (gic_handle_irq+0x6c/0xc4) [<c0008774>] (gic_handle_irq+0x6c/0xc4) from [<c0699404>] (__irq_svc+0x44/0x78) Exception stack(0xee061cd8 to 0xee061d20) 1cc0: 00000001 eebda7c8 1ce0: 00000000 eebda400 20000013 c12b5430 ec96c168 c12b542c c12b5430 00000001 1d00: 20000013 ee061e14 3eb13eb1 ee061d20 c00ba520 c069918c 20000013 ffffffff [<c0699404>] (__irq_svc+0x44/0x78) from [<c069918c>] (_raw_spin_unlock_irqrestore+0x40/0x68) [<c069918c>] (_raw_spin_unlock_irqrestore+0x40/0x68) from [<c0279f2c>] (__debug_object_init+0x30c/0x344) [<c0279f2c>] (__debug_object_init+0x30c/0x344) from [<c007b068>] (init_timer_on_stack_key+0x18/0x30) [<c007b068>] (init_timer_on_stack_key+0x18/0x30) from [<c0695248>] (schedule_timeout+0x8c/0x4c0) [<c0695248>] (schedule_timeout+0x8c/0x4c0) from [<c069768c>] (wait_for_common+0xec/0x164) [<c069768c>] (wait_for_common+0xec/0x164) from [<c0349fc4>] (msm_xfer_msg+0x3cc/0x51c) [<c0349fc4>] (msm_xfer_msg+0x3cc/0x51c) from [<c034a358>] (slim_sat_rxprocess+0x244/0x664) [<c034a358>] (slim_sat_rxprocess+0x244/0x664) from [<c0087290>] (process_one_work+0x354/0x648) [<c0087290>] (process_one_work+0x354/0x648) from [<c0089754>] (worker_thread+0x1a8/0x2a8) [<c0089754>] (worker_thread+0x1a8/0x2a8) from [<c008e480>] (kthread+0x90/0xa0) [<c008e480>] (kthread+0x90/0xa0) from [<c000f438>] (kernel_thread_exit+0x0/0x8) Make this lock irqsafe as well so that this potential bug doesn't occur. Change-Id: Icbef6d1d749ee6ee81b079e19e57f22c38f00c68 Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Aug 13, 2016
This moves ARM over to the asm-generic/unaligned.h header. This has the benefit of better code generated especially for ARMv7 on gcc 4.7+ compilers. As Arnd Bergmann, points out: The asm-generic version uses the "struct" version for native-endian unaligned access and the "byteshift" version for the opposite endianess. The current ARM version however uses the "byteshift" implementation for both. Thanks to Nicolas Pitre for the excellent analysis: Test case: int foo (int *x) { return get_unaligned(x); } long long bar (long long *x) { return get_unaligned(x); } With the current ARM version: foo: ldrb r3, [r0, stratosk#2] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 2B], MEM[(const u8 *)x_1(D) + 2B] ldrb r1, [r0, stratosk#1] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 1B], MEM[(const u8 *)x_1(D) + 1B] ldrb r2, [r0, #0] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D)], MEM[(const u8 *)x_1(D)] mov r3, r3, asl #16 @ tmp154, MEM[(const u8 *)x_1(D) + 2B], ldrb r0, [r0, stratosk#3] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 3B], MEM[(const u8 *)x_1(D) + 3B] orr r3, r3, r1, asl aosp-mirror#8 @, tmp155, tmp154, MEM[(const u8 *)x_1(D) + 1B], orr r3, r3, r2 @ tmp157, tmp155, MEM[(const u8 *)x_1(D)] orr r0, r3, r0, asl #24 @,, tmp157, MEM[(const u8 *)x_1(D) + 3B], bx lr @ bar: stmfd sp!, {r4, r5, r6, r7} @, mov r2, #0 @ tmp184, ldrb r5, [r0, aosp-mirror#6] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 6B], MEM[(const u8 *)x_1(D) + 6B] ldrb r4, [r0, aosp-mirror#5] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 5B], MEM[(const u8 *)x_1(D) + 5B] ldrb ip, [r0, stratosk#2] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 2B], MEM[(const u8 *)x_1(D) + 2B] ldrb r1, [r0, aosp-mirror#4] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 4B], MEM[(const u8 *)x_1(D) + 4B] mov r5, r5, asl #16 @ tmp175, MEM[(const u8 *)x_1(D) + 6B], ldrb r7, [r0, stratosk#1] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 1B], MEM[(const u8 *)x_1(D) + 1B] orr r5, r5, r4, asl aosp-mirror#8 @, tmp176, tmp175, MEM[(const u8 *)x_1(D) + 5B], ldrb r6, [r0, aosp-mirror#7] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 7B], MEM[(const u8 *)x_1(D) + 7B] orr r5, r5, r1 @ tmp178, tmp176, MEM[(const u8 *)x_1(D) + 4B] ldrb r4, [r0, #0] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D)], MEM[(const u8 *)x_1(D)] mov ip, ip, asl #16 @ tmp188, MEM[(const u8 *)x_1(D) + 2B], ldrb r1, [r0, stratosk#3] @ zero_extendqisi2 @ MEM[(const u8 *)x_1(D) + 3B], MEM[(const u8 *)x_1(D) + 3B] orr ip, ip, r7, asl aosp-mirror#8 @, tmp189, tmp188, MEM[(const u8 *)x_1(D) + 1B], orr r3, r5, r6, asl #24 @,, tmp178, MEM[(const u8 *)x_1(D) + 7B], orr ip, ip, r4 @ tmp191, tmp189, MEM[(const u8 *)x_1(D)] orr ip, ip, r1, asl #24 @, tmp194, tmp191, MEM[(const u8 *)x_1(D) + 3B], mov r1, r3 @, orr r0, r2, ip @ tmp171, tmp184, tmp194 ldmfd sp!, {r4, r5, r6, r7} bx lr In both cases the code is slightly suboptimal. One may wonder why wasting r2 with the constant 0 in the second case for example. And all the mov's could be folded in subsequent orr's, etc. Now with the asm-generic version: foo: ldr r0, [r0, #0] @ unaligned @,* x bx lr @ bar: mov r3, r0 @ x, x ldr r0, [r0, #0] @ unaligned @,* x ldr r1, [r3, aosp-mirror#4] @ unaligned @, bx lr @ This is way better of course, but only because this was compiled for ARMv7. In this case the compiler knows that the hardware can do unaligned word access. This isn't that obvious for foo(), but if we remove the get_unaligned() from bar as follows: long long bar (long long *x) {return *x; } then the resulting code is: bar: ldmia r0, {r0, r1} @ x,, bx lr @ So this proves that the presumed aligned vs unaligned cases does have influence on the instructions the compiler may use and that the above unaligned code results are not just an accident. Still... this isn't fully conclusive without at least looking at the resulting assembly fron a pre ARMv6 compilation. Let's see with an ARMv5 target: foo: ldrb r3, [r0, #0] @ zero_extendqisi2 @ tmp139,* x ldrb r1, [r0, stratosk#1] @ zero_extendqisi2 @ tmp140, ldrb r2, [r0, stratosk#2] @ zero_extendqisi2 @ tmp143, ldrb r0, [r0, stratosk#3] @ zero_extendqisi2 @ tmp146, orr r3, r3, r1, asl aosp-mirror#8 @, tmp142, tmp139, tmp140, orr r3, r3, r2, asl #16 @, tmp145, tmp142, tmp143, orr r0, r3, r0, asl #24 @,, tmp145, tmp146, bx lr @ bar: stmfd sp!, {r4, r5, r6, r7} @, ldrb r2, [r0, #0] @ zero_extendqisi2 @ tmp139,* x ldrb r7, [r0, stratosk#1] @ zero_extendqisi2 @ tmp140, ldrb r3, [r0, aosp-mirror#4] @ zero_extendqisi2 @ tmp149, ldrb r6, [r0, aosp-mirror#5] @ zero_extendqisi2 @ tmp150, ldrb r5, [r0, stratosk#2] @ zero_extendqisi2 @ tmp143, ldrb r4, [r0, aosp-mirror#6] @ zero_extendqisi2 @ tmp153, ldrb r1, [r0, aosp-mirror#7] @ zero_extendqisi2 @ tmp156, ldrb ip, [r0, stratosk#3] @ zero_extendqisi2 @ tmp146, orr r2, r2, r7, asl aosp-mirror#8 @, tmp142, tmp139, tmp140, orr r3, r3, r6, asl aosp-mirror#8 @, tmp152, tmp149, tmp150, orr r2, r2, r5, asl #16 @, tmp145, tmp142, tmp143, orr r3, r3, r4, asl #16 @, tmp155, tmp152, tmp153, orr r0, r2, ip, asl #24 @,, tmp145, tmp146, orr r1, r3, r1, asl #24 @,, tmp155, tmp156, ldmfd sp!, {r4, r5, r6, r7} bx lr Compared to the initial results, this is really nicely optimized and I couldn't do much better if I were to hand code it myself. Signed-off-by: Rob Herring <rob.herring@calxeda.com> Reviewed-by: Nicolas Pitre <nico@linaro.org> Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> modified for Mako from kernel.org reference Signed-off-by: faux123 <reioux@gmail.com>
spezi77
pushed a commit
to spezi77/kernel_msm
that referenced
this pull request
Aug 15, 2016
When sched_show_task() is invoked from try_to_freeze_tasks(), there is no RCU read-side critical section, resulting in the following splat: [ 125.780730] =============================== [ 125.780766] [ INFO: suspicious RCU usage. ] [ 125.780804] 3.7.0-rc3+ #988 Not tainted [ 125.780838] ------------------------------- [ 125.780875] /home/rafael/src/linux/kernel/sched/core.c:4497 suspicious rcu_dereference_check() usage! [ 125.780946] [ 125.780946] other info that might help us debug this: [ 125.780946] [ 125.781031] [ 125.781031] rcu_scheduler_active = 1, debug_locks = 0 [ 125.781087] 4 locks held by s2ram/4211: [ 125.781120] #0: (&buffer->mutex){+.+.+.}, at: [<ffffffff811e2acf>] sysfs_write_file+0x3f/0x160 [ 125.781233] stratosk#1: (s_active#94){.+.+.+}, at: [<ffffffff811e2b58>] sysfs_write_file+0xc8/0x160 [ 125.781339] stratosk#2: (pm_mutex){+.+.+.}, at: [<ffffffff81090a81>] pm_suspend+0x81/0x230 [ 125.781439] stratosk#3: (tasklist_lock){.?.?..}, at: [<ffffffff8108feed>] try_to_freeze_tasks+0x2cd/0x3f0 [ 125.781543] [ 125.781543] stack backtrace: [ 125.781584] Pid: 4211, comm: s2ram Not tainted 3.7.0-rc3+ #988 [ 125.781632] Call Trace: [ 125.781662] [<ffffffff810a3c73>] lockdep_rcu_suspicious+0x103/0x140 [ 125.781719] [<ffffffff8107cf21>] sched_show_task+0x121/0x180 [ 125.781770] [<ffffffff8108ffb4>] try_to_freeze_tasks+0x394/0x3f0 [ 125.781823] [<ffffffff810903b5>] freeze_kernel_threads+0x25/0x80 [ 125.781876] [<ffffffff81090b65>] pm_suspend+0x165/0x230 [ 125.781924] [<ffffffff8108fa29>] state_store+0x99/0x100 [ 125.781975] [<ffffffff812f5867>] kobj_attr_store+0x17/0x20 [ 125.782038] [<ffffffff811e2b71>] sysfs_write_file+0xe1/0x160 [ 125.782091] [<ffffffff811667a6>] vfs_write+0xc6/0x180 [ 125.782138] [<ffffffff81166ada>] sys_write+0x5a/0xa0 [ 125.782185] [<ffffffff812ff6ae>] ? trace_hardirqs_on_thunk+0x3a/0x3f [ 125.782242] [<ffffffff81669dd2>] system_call_fastpath+0x16/0x1b This commit therefore adds the needed RCU read-side critical section. Reported-by: "Rafael J. Wysocki" <rjw@sisk.pl> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
stratosk
pushed a commit
that referenced
this pull request
Sep 7, 2016
commit 8e96a87 upstream. Userspace can quite legitimately perform an exec() syscall with a suspended transaction. exec() does not return to the old process, rather it load a new one and starts that, the expectation therefore is that the new process starts not in a transaction. Currently exec() is not treated any differently to any other syscall which creates problems. Firstly it could allow a new process to start with a suspended transaction for a binary that no longer exists. This means that the checkpointed state won't be valid and if the suspended transaction were ever to be resumed and subsequently aborted (a possibility which is exceedingly likely as exec()ing will likely doom the transaction) the new process will jump to invalid state. Secondly the incorrect attempt to keep the transactional state while still zeroing state for the new process creates at least two TM Bad Things. The first triggers on the rfid to return to userspace as start_thread() has given the new process a 'clean' MSR but the suspend will still be set in the hardware MSR. The second TM Bad Thing triggers in __switch_to() as the processor is still transactionally suspended but __switch_to() wants to zero the TM sprs for the new process. This is an example of the outcome of calling exec() with a suspended transaction. Note the first 700 is likely the first TM bad thing decsribed earlier only the kernel can't report it as we've loaded userspace registers. c000000000009980 is the rfid in fast_exception_return() Bad kernel stack pointer 3fffcfa1a370 at c000000000009980 Oops: Bad kernel stack pointer, sig: 6 [#1] CPU: 0 PID: 2006 Comm: tm-execed Not tainted NIP: c000000000009980 LR: 0000000000000000 CTR: 0000000000000000 REGS: c00000003ffefd40 TRAP: 0700 Not tainted MSR: 8000000300201031 <SF,ME,IR,DR,LE,TM[SE]> CR: 00000000 XER: 00000000 CFAR: c0000000000098b4 SOFTE: 0 PACATMSCRATCH: b00000010000d033 GPR00: 0000000000000000 00003fffcfa1a370 0000000000000000 0000000000000000 GPR04: 0000000000000000 0000000000000000 0000000000000000 0000000000000000 GPR08: 0000000000000000 0000000000000000 0000000000000000 0000000000000000 GPR12: 00003fff966611c0 0000000000000000 0000000000000000 0000000000000000 NIP [c000000000009980] fast_exception_return+0xb0/0xb8 LR [0000000000000000] (null) Call Trace: Instruction dump: f84d0278 e9a100d8 7c7b03a6 e84101a0 7c4ff120 e8410170 7c5a03a6 e8010070 e8410080 e8610088 e8810090 e8210078 <4c000024> 48000000 e8610178 88ed023b Kernel BUG at c000000000043e80 [verbose debug info unavailable] Unexpected TM Bad Thing exception at c000000000043e80 (msr 0x201033) Oops: Unrecoverable exception, sig: 6 [#2] CPU: 0 PID: 2006 Comm: tm-execed Tainted: G D task: c0000000fbea6d80 ti: c00000003ffec000 task.ti: c0000000fb7ec000 NIP: c000000000043e80 LR: c000000000015a24 CTR: 0000000000000000 REGS: c00000003ffef7e0 TRAP: 0700 Tainted: G D MSR: 8000000300201033 <SF,ME,IR,DR,RI,LE,TM[SE]> CR: 28002828 XER: 00000000 CFAR: c000000000015a20 SOFTE: 0 PACATMSCRATCH: b00000010000d033 GPR00: 0000000000000000 c00000003ffefa60 c000000000db5500 c0000000fbead000 GPR04: 8000000300001033 2222222222222222 2222222222222222 00000000ff160000 GPR08: 0000000000000000 800000010000d033 c0000000fb7e3ea0 c00000000fe00004 GPR12: 0000000000002200 c00000000fe00000 0000000000000000 0000000000000000 GPR16: 0000000000000000 0000000000000000 0000000000000000 0000000000000000 GPR20: 0000000000000000 0000000000000000 c0000000fbea7410 00000000ff160000 GPR24: c0000000ffe1f600 c0000000fbea8700 c0000000fbea8700 c0000000fbead000 GPR28: c000000000e20198 c0000000fbea6d80 c0000000fbeab680 c0000000fbea6d80 NIP [c000000000043e80] tm_restore_sprs+0xc/0x1c LR [c000000000015a24] __switch_to+0x1f4/0x420 Call Trace: Instruction dump: 7c800164 4e800020 7c0022a6 f80304a8 7c0222a6 f80304b0 7c0122a6 f80304b8 4e800020 e80304a8 7c0023a6 e80304b0 <7c0223a6> e80304b8 7c0123a6 4e800020 This fixes CVE-2016-5828. Fixes: bc2a940 ("powerpc: Hook in new transactional memory code") Cc: stable@vger.kernel.org # v3.9+ Signed-off-by: Cyril Bur <cyrilbur@gmail.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Willy Tarreau <w@1wt.eu>
stratosk
pushed a commit
that referenced
this pull request
Jun 8, 2018
The logic in __ip6_append_data() assumes that the MTU is at least large enough for the headers. A device's MTU may be adjusted after being added while sendmsg() is processing data, resulting in __ip6_append_data() seeing any MTU. For an mtu smaller than the size of the fragmentation header, the math results in a negative 'maxfraglen', which causes problems when refragmenting any previous skb in the skb_write_queue, leaving it possibly malformed. Instead sendmsg returns EINVAL when the mtu is calculated to be less than IPV6_MIN_MTU. Found by syzkaller: kernel BUG at ./include/linux/skbuff.h:2064! invalid opcode: 0000 [#1] SMP KASAN Dumping ftrace buffer: (ftrace buffer empty) Modules linked in: CPU: 1 PID: 14216 Comm: syz-executor5 Not tainted 4.13.0-rc4+ #2 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 task: ffff8801d0b68580 task.stack: ffff8801ac6b8000 RIP: 0010:__skb_pull include/linux/skbuff.h:2064 [inline] RIP: 0010:__ip6_make_skb+0x18cf/0x1f70 net/ipv6/ip6_output.c:1617 RSP: 0018:ffff8801ac6bf570 EFLAGS: 00010216 RAX: 0000000000010000 RBX: 0000000000000028 RCX: ffffc90003cce000 RDX: 00000000000001b8 RSI: ffffffff839df06f RDI: ffff8801d9478ca0 RBP: ffff8801ac6bf780 R08: ffff8801cc3f1dbc R09: 0000000000000000 R10: ffff8801ac6bf7a0 R11: 43cb4b7b1948a9e7 R12: ffff8801cc3f1dc8 R13: ffff8801cc3f1d40 R14: 0000000000001036 R15: dffffc0000000000 FS: 00007f43d740c700(0000) GS:ffff8801dc100000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007f7834984000 CR3: 00000001d79b9000 CR4: 00000000001406e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: ip6_finish_skb include/net/ipv6.h:911 [inline] udp_v6_push_pending_frames+0x255/0x390 net/ipv6/udp.c:1093 udpv6_sendmsg+0x280d/0x31a0 net/ipv6/udp.c:1363 inet_sendmsg+0x11f/0x5e0 net/ipv4/af_inet.c:762 sock_sendmsg_nosec net/socket.c:633 [inline] sock_sendmsg+0xca/0x110 net/socket.c:643 SYSC_sendto+0x352/0x5a0 net/socket.c:1750 SyS_sendto+0x40/0x50 net/socket.c:1718 entry_SYSCALL_64_fastpath+0x1f/0xbe RIP: 0033:0x4512e9 RSP: 002b:00007f43d740bc08 EFLAGS: 00000216 ORIG_RAX: 000000000000002c RAX: ffffffffffffffda RBX: 00000000007180a8 RCX: 00000000004512e9 RDX: 000000000000002e RSI: 0000000020d08000 RDI: 0000000000000005 RBP: 0000000000000086 R08: 00000000209c1000 R09: 000000000000001c R10: 0000000000040800 R11: 0000000000000216 R12: 00000000004b9c69 R13: 00000000ffffffff R14: 0000000000000005 R15: 00000000202c2000 Code: 9e 01 fe e9 c5 e8 ff ff e8 7f 9e 01 fe e9 4a ea ff ff 48 89 f7 e8 52 9e 01 fe e9 aa eb ff ff e8 a8 b6 cf fd 0f 0b e8 a1 b6 cf fd <0f> 0b 49 8d 45 78 4d 8d 45 7c 48 89 85 78 fe ff ff 49 8d 85 ba RIP: __skb_pull include/linux/skbuff.h:2064 [inline] RSP: ffff8801ac6bf570 RIP: __ip6_make_skb+0x18cf/0x1f70 net/ipv6/ip6_output.c:1617 RSP: ffff8801ac6bf570 Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Mike Maloney <maloney@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> (cherry picked from commit 749439b) Bug: 65023306 Change-Id: I3b713621c749b7fd3a070116be8996ae2e2dd6e8 Signed-off-by: Greg Hackmann <ghackmann@google.com>
stratosk
pushed a commit
that referenced
this pull request
Jun 8, 2018
syzkaller with KASAN reported an out-of-bounds read in asn1_ber_decoder(). It can be reproduced by the following command, assuming CONFIG_X509_CERTIFICATE_PARSER=y and CONFIG_KASAN=y: keyctl add asymmetric desc $'\x30\x30' @s The bug is that the length of an ASN.1 data value isn't validated in the case where it is encoded using the short form, causing the decoder to read past the end of the input buffer. Fix it by validating the length. The bug report was: BUG: KASAN: slab-out-of-bounds in asn1_ber_decoder+0x10cb/0x1730 lib/asn1_decoder.c:233 Read of size 1 at addr ffff88003cccfa02 by task syz-executor0/6818 CPU: 1 PID: 6818 Comm: syz-executor0 Not tainted 4.14.0-rc7-00008-g5f479447d983 #2 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:16 [inline] dump_stack+0xb3/0x10b lib/dump_stack.c:52 print_address_description+0x79/0x2a0 mm/kasan/report.c:252 kasan_report_error mm/kasan/report.c:351 [inline] kasan_report+0x236/0x340 mm/kasan/report.c:409 __asan_report_load1_noabort+0x14/0x20 mm/kasan/report.c:427 asn1_ber_decoder+0x10cb/0x1730 lib/asn1_decoder.c:233 x509_cert_parse+0x1db/0x650 crypto/asymmetric_keys/x509_cert_parser.c:89 x509_key_preparse+0x64/0x7a0 crypto/asymmetric_keys/x509_public_key.c:174 asymmetric_key_preparse+0xcb/0x1a0 crypto/asymmetric_keys/asymmetric_type.c:388 key_create_or_update+0x347/0xb20 security/keys/key.c:855 SYSC_add_key security/keys/keyctl.c:122 [inline] SyS_add_key+0x1cd/0x340 security/keys/keyctl.c:62 entry_SYSCALL_64_fastpath+0x1f/0xbe RIP: 0033:0x447c89 RSP: 002b:00007fca7a5d3bd8 EFLAGS: 00000246 ORIG_RAX: 00000000000000f8 RAX: ffffffffffffffda RBX: 00007fca7a5d46cc RCX: 0000000000447c89 RDX: 0000000020006f4a RSI: 0000000020006000 RDI: 0000000020001ff5 RBP: 0000000000000046 R08: fffffffffffffffd R09: 0000000000000000 R10: 0000000000000002 R11: 0000000000000246 R12: 0000000000000000 R13: 0000000000000000 R14: 00007fca7a5d49c0 R15: 00007fca7a5d4700 Fixes: 42d5ec2 ("X.509: Add an ASN.1 decoder") Cc: <stable@vger.kernel.org> # v3.7+ Signed-off-by: Eric Biggers <ebiggers@google.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: James Morris <james.l.morris@oracle.com> (cherry picked from commit 2eb9eab) Bug: 73827422 Change-Id: I3c57fd16ebc63214c4225e85004b1339fbc41728 Signed-off-by: Connor O'Brien <connoro@google.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I patched with CM 11 kind of "colorfixing" so that Semaphore could be used on CM 11 variances.
Also I patched with kexec hardboot so that it could be used with Multiboot TWRP by Tassadar
Also, Faux 3.2 driver was included